Analysis of Alaska Transportation Sectors to Assess Energy Use and Impacts of Price Shocks and Climate Change Legislation

Home , Economy of Alaska, Port of Anchorage, Southeast Alaska

Item Type Report

Authors Fay, Ginny; Schwörer, Tobias; Guettabi, Mouhcine; Armagost, Jeffrey

Publisher Alaska University Transportation Center, UAA Institute of Social and Economic Research

Download date 30/09/2021 10:59:12

Link to Item http://hdl.handle.net/11122/7467

Analysis of Alaska Transportation Sectors to Assess Energy Use

and Impacts of Price Shocks and Climate Change Legislation Alaska University Transportation University Alaska Center

and Social ResearchEconomic of UAA Institute

Photo

Authors: Ginny Fay, Tobias Schwörer, Mouhcine Guettabi, Jeffrey Armagost

Date:

April 2013

Prepared By: Institute of Social and Economic Research, University of Alaska Anchorage

Alaska University Transportation Center UAA Institute of Social and Economic Duckering Building Room 245 Research P.O. Box 755900 3211 Providence Dr. Fairbanks, AK 99775-5900 Anchorage, AK 99508

INE/AUTC13.03

Form approved OMB No. REPORT DOCUMENTATION PAGE

Public reporting for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestion for reducing this burden to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-1833), Washington, DC 20503 1. AGENCY USE ONLY (LEAVE 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED BLANK) April 2013 Final Report (8/1/2009-4/31/2013)

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Analysis of Alaska Transportation Sectors to Assess Energy Use and Impacts of Price Shocks and Climate Change Legislation 309002 DTRT06-G-0011 6. AUTHOR(S) Ginny Fay, Tobias Schwörer, Mouhcine Guettabi, Jeffrey Armagost

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION Institute of Social and Economic Research REPORT NUMBER University of Alaska Anchorage 3211 Providence Drive Anchorage, Alaska 99508

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING Alaska University Transportation Center AGENCY REPORT NUMBER PO Box 755900 INE/AUTC13.03 Fairbanks, AK 99775-5900

11. SUPPLENMENTARY NOTES

12a. DISTRIBUTION / AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE No restrictions

13. ABSTRACT (Maximum 200 words) We analyzed the use of energy by Alaska’s transportation sectors to assess the impact of sudden fuel prices changes. We conducted three types of analysis: 1) Development of broad energy use statistics for each transportation sector, including total annual energy and fuel use, carbon emissions, fuel use per ton-mile and passenger-mile, and cost of fuel per ton-mile and passenger-mile. 2) Economic input-output analysis of air, rail, truck, and water transportation sectors. 3) Adjustment of input-output modeling to reflect sudden fuel price changes to estimate the potential impact on industry output and employment. Alaska air transportation used approximately 1.9 billion gallons of fuel annually; 961 million gallons were used for intra-state and exiting Alaska flights. Water transportation used 101.8 million gallons annually, approximately 84.3 million gallons for intra-state and exiting segments. Railroad and truck transportation used 5.1 and 8.8 million gallons annually, respectively. Simulated fuel price increases resulted in an estimated $456.8 million in value-added losses to the Alaska economy through the increase in cost of transportation services, as well as an equivalent loss in income to Alaska household of $26.8 million. A carbon emissions tax would have the greatest impact on the cost of air transportation services followed by water, trucking and rail.

14- KEYWORDS : transportation, fuel prices, emissions, Alaska, air transportation, water transportation, rail 15. NUMBER OF PAGES transportation, truck transportation, energy economics 16. PRICE CODE N/A 17. SECURITY CLASSIFICATION 18. SECURITY 19. SECURITY 20. LIMITATION OF OF REPORT CLASSIFICATION OF THIS CLASSIFICATION OF ABSTRACT Unclassified PAGE ABSTRACT N/A Unclassified Unclassified NSN 7540-01-280-5500 STANDARD FORM 298 (Rev. 2-98) Prescribed by ANSI Std. 239-18 298

Contents List of Figures ...... iii List of Tables ...... iv Acknowledgments ...... vi Abstract ...... vii Executive Summary ...... 1 Introduction ...... 6 Background ...... 7 Alaska’s Transportation Modes ...... 9 Water Transportation ...... 9 Trucking ...... 13 Railroads...... 15 Aviation ...... 16 Research Approach ...... 17 Water Transportation ...... 18 Marine Ships ...... 18 Barges ...... 19 Ferries ...... 19 Land Transportation ...... 20 Railroad ...... 20 Trucking ...... 20 Aviation ...... 20 Economic Impact Analysis of Fuel Price Changes on the Alaska Economy ...... 21 IMPLAN ...... 22 Findings and Applications ...... 23 Fuel Use and Efficiency by Mode ...... 23 Water Transportation ...... 23 Land Transportation ...... 30 Aviation ...... 33 Intermodal Fuel Use Comparisons ...... 35 Change in Transportation Costs and Impact on Alaska Industries ...... 40

Impact of Change in Transportation Costs on Alaska Households ...... 52 Carbon Emissions Analysis ...... 55 Direct Emissions from Fuel Use ...... 55 Carbon Emissions Tax Analysis ...... 57 Conclusions ...... 61 References ...... 64 Appendix A. Marine Transportation Companies ...... A-1 Marine Ships ...... A-1 Barges ...... A-1 Appendix B. Barge Fuel Use Calculations ...... B-1 Scheduled Barge Traffic ...... B-1 Unscheduled Barge Traffic ...... B-2 Liquid Barge Traffic ...... B-2 Final Calculations ...... B-3 Limitation of the Analysis ...... B-5 Appendix C. Data Dictionary of Variables and Sources Used for Aviation Fuel Estimates ...... C-1 Appendix D. Glossary of Economic Impact Terms ...... D-1

List of Figures Figure ES1. Comparison of fuel use and costs per ton-mile for Alaska transportation, 2007–2010, 2011$ ...... 2 Figure ES2. Comparison of fuel use and costs per passenger-mile for rail and air, 2007–2010, 2011$ ...... 3 Figure ES3. Comparison annual emissions by transportation sector, intra-state and exiting only, 2007–2010 ...... 4 Figure 1. Alaska transportation system ...... 8 Figure 2. Passenger delivery in Grayling (Photo credit: S.G. Colt) ...... 8 Figure 3. Estimate of imports to Alaska via water transportation, thousands of short tons, 2010 ...... 10 Figure 4. Estimate of exports from Alaska via water transportation, thousands of short tons, 2010 ...... 11 Figure 5. Alaska Marine Highway System communities and routes ...... 12 Figure 6. Inter-Island Ferry Authority communities and routes ...... 13 Figure 7. Map of Alaska major roadways ...... 14 Figure 8. Alaska Railroad route map ...... 16 Figure 9. Annual crude oil prices, 2000–2012 ...... 23 Figure 10. Alaska Railroad freight transport in short tons, petroleum and all other goods (AOG), 2007–2009 ...... 31 Figure 11. Comparison of annual fuel use, 2007–2010 ...... 35 Figure 12. Comparison of annual fuel costs, 2007–2010, 2011$ ...... 36 Figure 13. Comparison of annual fuel prices per gallon, 2007–2010, 2011$ ...... 36 Figure 14. Comparison of fuel use and costs per ton-mile for Alaska transportation, 2007–2010, 2011$ ...... 39 Figure 15. Comparison of fuel use and costs per passenger-mile for rail and air, 2007–2010, 2011$ ...... 39 Figure 16. Comparison of annual emissions by transportation sector, intrastate and exiting only, 2007–2010 ...... 55

iii

List of Tables Table 1. Long-haul trucking companies providing services in Alaska ...... 14 Table 2. Illustrative input-output transactions table (in millions of dollars) ...... 22 Table 3. Estimated fuel usage and costs for freight movement by marine vessels 2006–2010, 2011$ ...... 25 Table 4. Estimated tons, fuel usage and costs for freight movement by barge 2006–2012, 2011$ ...... 25 Table 5. Estimated regional fuel usage for freight movement by barge 2006–2012, 2011$ ...... 26 Table 6. Passengers, freight, fuel usage, and costs for the Alaska Marine Highway System and Inter-Island Ferry Authority 2007–2010, 2011$ ...... 28 Table 7. Estimated tons, fuel usage, and costs for freight movement for the Alaska Marine Highway System and Inter-Island Ferry Authority 2007–2010, 2011$ ...... 29 Table 8. Estimated tons, fuel usage, and costs for passenger movement by the Alaska Marine Highway System and Inter-Island Ferry Authority, 2007–2010, 2011$ ...... 29 Table 9. Annual passengers and freight by segment type for the Alaska Marine Highway System and Inter-Island Ferry Authority segment types 2007–2010, 2011$ ...... 30 Table 10. Fuel usage and costs for freight movement by the Alaska Railroad 2007–2010, 2011$ ...... 31 Table 11. Alaska Railroad freight transport in short tons, petroleum and all other goods ...... 31 Table 12. Average fuel usage and costs for freight movement by the Alaska Railroad 2007–2010, 2011$ ...... 32 Table 13. Fuel usage and costs for passenger movement by the Alaska Railroad 2007–2010, 2011$ ...... 32 Table 14. Average fuel usage and costs for passenger movement by the Alaska Railroad 2007– 2010, 2011$ ...... 32 Table 15. Estimated fuel usage and costs by long-haul trucks, 2011$ ...... 33 Table 16. Fuel usage and costs for intra-state scheduled air transportation in Alaska 2005–2010, 2011$ ...... 34 Table 17. Estimated transportation fuel use and costs, Alaska 2007–2010 ...... 37 Table 18. Comparison of fuel use and costs per ton-mile for Alaska water transportation, 2007– 2010, 2011$ ...... 38 Table 19. Comparison of fuel use and costs per ton-mile for rail and trucking, 2007–2010, 2011$ ...... 38 Table 20. Comparison of fuel use and costs per passenger-mile for rail, ferry and air, 2007–2010, 2011$ ...... 38 Table 21. Alaska transportation sector employment, payroll, and firms by transportation mode, 2008 ...... 40 Table 22. Alaska transportation sector employment, payroll, and firms by transportation mode, 2010 ...... 41 Table 23. Change in Alaska transport sector employment, payroll, and firms, 2008–2010...... 42 Table 24. Number of transportation industry firms by size, 2008* ...... 43

Table 25. Number of transportation industry firms by size, 2010* ...... 44 Table 26. Alaska transportation services estimated price increase, 2008–2010 ...... 45 Table 27. Transportation usage intensity, 2008 ...... 46 Table 28. Value-added losses (most affected industries in absolute terms), millions$$ ...... 47 Table 29. Value-added losses (most affected industries relative to own value added) ...... 48 Table 30. Change in institution and industry commodity demand for transportation services between 2008 and 2010 ...... 49 Table 31. Alaska industries with largest transportation expenditure inputs, 2008, million$$...... 50 Table 32. Mode shifts for Alaska industries with largest transportation expenditure inputs following imposed fuel price increases ...... 51 Table 33. Institution and industry demand for water transportation, 2008 and 2010 (2008$) ...... 52 Table 34. Increases in the cost of transportation services to households by income groups after an increase in the price of refined petroleum products, millions$$ ...... 52 Table 35. Increases in the cost of air, rail, water, and truck transportation services to households by income groups after an increase in the price of refined petroleum products, millions$$ ...... 53 Table 36. Shift in household purchases of air, rail, water, and truck transportation services due to increases in the cost of transportation services ...... 54 Table 37. Increases in household expenditures due to an increase in the cost of refined fuel prices, millions$$ ...... 54 Table 38. Estimated aviation, shipping, trucking, and rail emissions, Alaska 2007–2010 ...... 56 Table 39. Estimated water transportation ton-miles per gallon and fuel costs and emissions per ton-mile transported, Alaska 2007–2010, 2011$ ...... 57 Table 40. Comparison of estimated land transportation modes ton-miles per gallon and fuel costs and emissions per ton-mile transported, Alaska 2007–2010, 2011$ ...... 57 Table 41. Comparison of estimated passenger transportation modes passenger-miles per gallon and fuel costs and emissions per passenger-mile transported, Alaska 2007–2010, 2011$ ...... 57 Table 42. Industry sectors most impacted by a potential carbon emissions tax ...... 58 Table 43. Transportation sector fuel use and emissions by industry output ...... 59 Table C1. Fuel usage and costs for intra-state scheduled air transportation in Alaska ...... C-5 Table C2. Fuel usage and costs for intra-state non-scheduled air transportation in Alaska ...... C-6 Table C3. Fuel usage and costs for exiting scheduled and non-scheduled air transportation in Alaska ...... C-7 Table C4. Fuel usage and costs for entering scheduled and non-scheduled air transportation in Alaska ...... C-8

Acknowledgments The research reported here was performed under a U.S. Department of Transportation, Research and Innovative Technology Administration (RITA) grant DTRT06-G-0011, provided through the Alaska University Transportation Center. Matching funds were provided from the University of Alaska Foundation’s BP/Conoco Phillips fund and were awarded by then-President Mark Hamilton to the Institute of Social and Economic Research (ISER), University of Alaska Anchorage, for energy-related programs. We sincerely appreciate these matching funds that made it possible to conduct this research.

Virginia “Ginny” Fay, assistant professor of economics at ISER, was the project director and principal investigator of this research. The other authors of this report are Dr. Mouhcine Guettabi, assistant professor of economics, ISER; and Tobias Schwörer and Jeffrey Armagost, research professionals, ISER. Dr. Stephen Colt, professor of economics, ISER, and Dr. Matthew Berman, professor of economics, ISER, both provided critical advice and review. Katherine Jackstadt, research professional, ISER, was instrumental in collecting and organizing data.

This project required extensive data collection and analysis, made possible through the cooperation and assistance of a number of Alaska state agencies and private companies, including the Alaska Department of Transportation and Public Facilities, especially Jeff Ottesen, director of Statewide Planning, Catherine Belfry with the Alaska Marine Highway System, Peter Freer, Statewide Planning, Clint Adler, Chief, Research Development and Technology Transfer, and Bruce Carr and Steve Silverstein, Alaska Railroad. Many people improved our understanding of Alaska shipping including Rick Kessler, Gunther Hoock, and Eddie Walton, Horizon Lines; Renata Benett and Carlos Roldan, TOTE; Paul Friese, Lynden Transport; Larry Stauffer, Northland Services; and Mark Smith and Justin Charon, Vitus Marine. Holly Baker-Kjostad, Carlile Transportation Systems, Inc., was tremendously helpful at explaining Alaska trucking as was Sean O'Hare, Alaska Traffic Company. Roberta Landgren, Inter-Island Ferry Authority, dedicated time to provide accurate data and assistance.

Suggested citation: Fay, Ginny, Tobias Schwörer, Mouhcine Guettabi, Jeffrey Armagost, 2013, Analysis of Alaska Transportation Sector to Assess Energy Use and Impacts of Price Shocks and Climate Change Legislation, Institute of Social and Economic Research, University of Alaska Anchorage, prepared for the Alaska University Transportation Center pp. 94.

Abstract This project analyzed the use of energy by Alaska’s transportation sectors to assess the impact of sudden fuel price changes. We conducted three primary types of analysis: (1) Development of broad energy use statistics for each transportation sector, such as estimated total annual energy and fuel use, carbon emissions, fuel use per ton-mile and passenger-mile, and cost of fuel per ton-mile and passenger-mile. (2) Economic input-output analysis, which estimates the employment and output of air, rail, truck, and water transportation sectors in the Alaska economy. (3) Adjustment of input-output modeling assumptions to reflect sudden fuel price changes and/or emissions taxes to estimate the potential impact of these changes on industry output and employment in the Alaska economy. We found that Alaska air transportation used approximately 1.9 billion gallons of fuel annually, of which 961 million gallons were used for intra-state and exiting Alaska flights. Water transportation (ships, barges, and ferries) used 101.8 million gallons of fuel annually, with approximately 84.3 million gallons for intra- state and exiting segments. Railroad transportation used 5.1 million gallons of fuel annually, and truck transportation used 8.8 million gallons of fuel annually. The impact of fuel price increases similar to those that occurred between 2008 and 2010 results in an estimated $456.8 million in value-added losses to the Alaska economy through cost increases of transportation services. The cost increases, or equivalent loss in income, to Alaska households are $26.8 million. A carbon emissions tax would have the greatest impact on the cost of air transportation services, followed by water, trucking, and rail.

vii

Executive Summary This project analyzed energy use by Alaska’s transportation sectors to assess the impact of sudden fuel price changes or carbon emissions taxes. Inexpensive fossil fuels helped nurture Alaska’s early economic growth. Over time, key Alaska industries such as fishing, mining, tourism, and transportation, as well as activities such as subsistence gathering, have grown to depend directly on liquid fossil fuels.

Compared with other states, Alaska is unique in its energy use. In 2010, for example, per capita energy consumption in Alaska was triple the national average. High energy use makes the Alaska economy more vulnerable to energy price volatilities and shocks. For state policy makers and industry, such vulnerability necessitates a better understanding of how energy prices and legislation affect transportation patterns and efficiency.

The relationship of transportation to greenhouse gas (GHG) emissions is also important in the context of ongoing social and political discussion of climate change. Transportation is a major contributor to the

GHG emissions (primarily carbon dioxide, CO2) associated with increased global temperatures; almost 30% of U.S. GHG emissions come from transportation. Additionally, transportation assets and operations worth billions of dollars are vulnerable to the impacts of climate change. Freight GHG is growing at a rate three times that of passenger GHG.

We conducted three primary types of analysis: (1) Development of broad energy use statistics for each transportation sector. We estimated the energy and fuel used by the air, water, trucking, and rail transportation sectors. We compared their fuel intensity to move passengers and freight by estimating their passenger-miles per gallon of fuel, ton-miles per gallon of fuel, fuel costs per passenger-mile and per ton-mile, and

CO2 emissions per ton-mile and passenger-mile. (2) Economic input-output analysis, to estimate the employment and output of air, rail, truck, and water transportation sectors in the Alaska economy. (3) Adjustment of input-output modeling assumptions to reflect sudden fuel price changes and/or emissions taxes (which function similarly to an increase in fuel prices) to estimate the potential impact of these changes on industry output, employment, and Alaska households.

We analyzed the impact of fuel price changes that occurred in 2008 and, by examining income, employment, and industry output changes in 2010, how the Alaska economy, in the long run, adjusted to higher prices and potential price volatility.

We estimated that rail is the most efficient form of transportation for moving freight per gallon of fuel, followed by barge, marine ship, truck, and ferry. In the lower 48 water transportation is consistently found to be most fuel efficient; Alaska water transportation may be less efficient than rail because of the less than full back-hauls over long distances. As measured by passenger-miles per gallon of fuel, we again found that rail transport is the most fuel-efficient, followed by air and ferry. Fuel costs per ton- mile and passenger-mile followed the same pattern of efficiency (Figures ES1 and 2), as well as CO2 emissions intensity (Figure ES3). Note that ferries provide essential transportation services to locations not connected by roads. Given the age and configurations of Alaska ferries, the policy of avoiding direct

competition with the private sector, and the schedules operated to meet the needs of the traveling public, it is unlikely that ferries could ever be as fuel-efficient as their private sector counterparts are.

Faced with continued high or increasing fuel prices or carbon legislation, the demand for Alaska Railroad transportation services could potentially increase with shifts away from trucking. However, because the distance from Anchorage to Fairbanks and the Kenai Peninsula are relatively short, freight handling would have to be quite efficient, and wages competitive, to compete with the comparative efficiency of truck transportation, with its fewer freight intermodal transfers.

Figure ES1. Comparison of fuel use and costs per ton-mile for Alaska transportation, 2007–2010, 2011$ (Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations)

*U.S. average for comparison purposes only. Figure ES2. Comparison of fuel use and costs per passenger-mile for rail and air, 2007–2010, 2011$ (Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Army Corps of Engineers, Waterborne Statistics; AMHS; AKRR; IFA; Ingram, 2008; company proprietary information; author calculations)

Millions CO2 tons Millions 4

- 2007 2008 2009 2010 Air Water Trucks Rail

Figure ES3. Comparison of annual emissions by transportation sector, intra-state and exiting only, 2007–2010 Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

To connect fuel price-related changes in transportation costs to impacts on the Alaska economy, we examined several factors, including industry and household transportation uses. Analysis of the Alaska economy found that the ten industries most dependent on transportation services are: 1. Seafood product preparation and packaging 2. Support activities for oil and gas operations 3. Transport by truck 4. Drilling of oil and gas wells 5. Construction of new nonresidential commercial and health care structures 6. Construction of new residential single-/multi-family housing 7. Electric power generation, transmission, distribution 8. Mining of gold, silver, and other metal ore 9. Food services and drinking places 10. Other state and local government enterprises

Consequently, these industries are the most affected by increases in fuel prices or other changes that raise the cost of transportation services as an input in their production. Most of these are core industries in the Alaska economy.

The ten Alaska industries that would be most affected by carbon emissions legislation are: 1. Petroleum refineries 2. Natural gas distribution 3. State and local government electric use 4. Asphalt paving mixture and block manufacturing 5. State and local government passenger 6. Other basic chemical manufacturing 7. Transport by pipeline 8. Plastics material and resin manufacturing 9. Commercial fishing 10. Transport by air

Because of the fuel and carbon intensity of air transportation, airlines have made a sustained effort to improve air transportation efficiency with increased load factors and increased fuel efficiency of airplanes. However, despite the increases in fuel efficiency, of the four transportation sectors analyzed, air transportation continues to be the most vulnerable to emissions legislation impacts.

These ten industries potentially most affected by carbon emissions legislation are also the industries where increased efficiencies and reduced dependence on fossil fuels could have the most payback, as measured by avoiding potential emissions tax impacts.

Alaska households at all income levels are also vulnerable to increases in the price of transportation services as a result of fuel price increases or carbon emissions legislation. Transportation services include the direct purchases of things like passenger tickets for air, rail, or ferry tickets as well as the embedded costs of transportation services in groceries or furniture. If Alaska households continued to

purchase transportation services at the same level after fuel price increases similar to those that occurred between 2008 and 2010, these services would cost an additional $26.8 million; an estimated 73% of these cost increases would be paid by households earning over $50,000 annually. In all likelihood, however, households would reduce their spending on transportation services. Water and truck transportation services declined the most in our simulation, probably because the majority of goods Alaska households routinely purchase are transported by water and truck.

In addition to the higher cost of transportation services resulting from higher fuel prices, direct purchases of refined petroleum products would cost Alaska households an additional $124.1 million if the households continued to purchase at the same level after fuel price increases similar to those that occurred between 2008 and 2010. Similar to price increases for transportation services, households with incomes of $50,000 or higher would absorb an estimated 70% of the refined petroleum price increases.

A recent National Cooperative Freight Research Program (NHCRP) report (Holguín-Veras, José, et al, 2013), found that there is a lack of freight cost data for the various modes of freight transportation, and that no single source can provide the key cost data for any mode. In most cases, some data are available in the reports published by public-sector agencies, trade groups, and research universities. However, because these data were collected in response to the needs of specific projects, they cannot replace data formally collected as part of regularly scheduled data collection efforts. Publicly available cost data for air freight and terminals are practically nonexistent. Unfortunately, no single data source could fill all the gaps in freight cost data. Our data collection efforts and analyses were hampered by this data issue.

Our economic impact simulation did not include utilities, so price increases for space heating and electricity are not included in these estimates.

Introduction Inexpensive fossil fuels helped drive Alaska’s early economic growth. Over time, key Alaska industries such as fishing, mining, tourism, and transportation, as well as activities such as subsistence gathering, have grown to depend directly on liquid fossil fuels including diesel, gasoline, and jet fuels. In addition, Alaska’s urban service economy has depended heavily on a relatively low cost of living, facilitated by low energy prices in Southcentral Alaska, and doing business has historically been assisted by cheap transportation fuels.

These conditions are changing rapidly and perhaps permanently. Although Alaska has a low absolute energy demand compared with the U.S. average, its per capita energy consumption is the highest in the country—more than three times the U.S. average (U.S. DOE, EIA, 2012a). A number of factors contribute to the state’s higher per capita energy consumption. Alaska’s role as a major world air cargo and transportation hub, oil producer, and marginal refiner substantially increases the per capita use calculation. Alaska’s remoteness and dispersed populations, along with a limited road system, cause Alaskans to depend more on air transportation services. The relatively greater dependence of Alaska industries and residents on energy creates a higher vulnerability to energy price volatilities and shocks. Such vulnerability means that state policy makers and industry need to better understand how energy prices and legislation affect transportation patterns and efficiency. Transportation is a major contributor to greenhouse gas (GHG) emissions—almost 30% of U.S. GHG emissions—associated with increased global temperatures, and transportation assets and operations worth billions of dollars are vulnerable to the impacts of climate change (TRB, 2012a, b). Freight transportation GHG is growing at a rate three times that of passenger transportation GHG (TRB, 2012b).

This project analyzed the use of energy by Alaska’s transportation sectors to assess what might happen if fuel prices suddenly change. We conducted three primary types of analysis:

(1) Development of broad energy use statistics for each transportation sector, such as estimated total annual energy and fuel use, carbon emissions, fuel use per ton-mile and passenger-mile, and cost of fuel per ton-mile and passenger-mile.

(2) Economic input-output analysis, which estimates the employment and output of air, rail, truck, and water transportation sectors in the Alaska economy.

(3) Adjustment of input-output modeling assumptions to reflect fuel price shocks/changes and/or emissions taxes to estimate the potential effect of these changes on industry output and employment in the Alaska economy.

We analyzed the impact of fuel price changes that occurred in 2008 and, by examining income, employment, and industry output changes in 2010, we analyzed how the Alaska economy adjusted to higher prices and potential price volatility in the long run.

The report is organized into background, research approach, and findings and applications sections, each of which is organized by transportation mode—water, rail, trucking, and air. The findings and

applications section is organized by transportation mode, fuel use, economic impact analysis, and emissions.

Background Given its geography, Alaska has long relied on aviation and marine transportation to move people and goods (Figure 1). Freight transport for goods used in Alaska continues to be dominated by marine transportation, as has been the case since Russian colonization (Gray and Rowe, 1982). Although Alaska is the largest state by area, its road mileage is the fifth lowest in the nation, leaving 82% of its communities unconnected to a state road system (Schultz, 2012). The reasons for Alaska’s limited road system are many, and the state’s unusual dependence on efficient intermodal transportation will no doubt continue. Extreme weather, rugged terrain, vast distances, low population density, and scattered islands make future road construction initiatives for connecting communities to the road system difficult and extremely costly when compared with the number of end users (ADOT&PF, 2008). Residents of these rural areas not connected to the state’s road system primarily use expensive air transportation for passenger and consumer goods movement.

In more populated areas, intermodal reliance looks quite different. More than half of the state’s population resides within the “Railbelt,” the region served by the Alaska Railroad (AKRR) and the state highway system. This region and a few small urban areas in Southeast Alaska have competing transportation modes, services, and economies of scale for freight and passengers. The major changes in Alaska’s transportation system in the last 50 years have primarily been technological improvements within each transportation mode rather than major system changes.

From an economist’s perspective, understanding Alaska’s highly intermodal transportation system requires a focus on inputs and outputs that are specific to that system, especially energy resources. The journey of freight goods to Alaska consumers offers a good illustration. Most of the food, household, and consumer goods shipped to Alaska from the continental United States begin their journey at manufacturing plants or distribution facilities. Trucks or trains then transport the goods to ports in either Tacoma or Seattle, Washington, where they are loaded onto container ships, barges, or roll-on/roll-off vessels for shipment to Alaska ports. If bound for a community connected to the highway system, the freight often completes its journey in trucks. Freight also travels north via the Alaska Railroad. Freight destined for towns off the road system is flown from either Anchorage or Fairbanks to remote communities and then is either driven by pickup truck if there is a regional road system or loaded onto smaller aircraft or boats for shipment to outlying villages. Quite often in remote areas, freight makes the final leg of the journey in sleds pulled by snow machines or on four-wheelers (ADOT&PF, 2008) (Figure 2). Each leg of this journey involves a specific mode and energy resource—nearly always liquid fuel.

Figure 1. Alaska transportation system

Figure 2. Passenger delivery in Grayling (Photo credit: S.G. Colt)

Alaska’s Transportation Modes Water Transportation Alaska depends more heavily on water transportation than does any other state in the continental U.S. Water transportation is one of the smaller transportation sectors as measured by employment, but it handles the greatest tonnage of freight entering Alaska. Access to navigable water has been a critical factor in Alaska’s development, often to the extent of dictating the location of communities. Even the Interior community of Fairbanks owes its existence to its location on the Chena River. At 33,900 miles, the shoreline of Alaska is far greater than that of the entire Lower 48. Commercial shippers serve this extensive coastline as far north as Prudhoe Bay. The Yukon, Tanana, and Kuskokwim Rivers and some of their tributaries also are important shipping routes for nearby communities (Fried and Keith, 2005).

Ports and harbors within coastal and riverine communities are an integral part of the freight transportation network. Ports are involved in the transport of forest products, oil and bulk petroleum, coal, seafood, general cargo, and consumer goods. While overland trucking and rail are important for delivery within the state, marine and air transport dominate Alaska‘s interstate freight movement (ADOT&PF, 2008). There are approximately 476 public and private ports and harbors in Alaska—240 in the southeast region and 236 in the southwest and western regions combined. This figure does not include barge landing and boat haul-out facilities along the riverine communities of the Kuskokwim and Yukon Rivers.

The Port of Anchorage (POA) is Alaska’s major port. Annual cargo entering the POA—most of it originating at the Port of Tacoma—accounts for an estimated 90% of the merchandise used by Alaska communities west of Cordova (UAA, 2011). Shipments bound for Alaska are nearly 30% of Tacoma’s total cargo activity. The value of these goods is estimated at well over $1 billion annually (Chase, 2004).

The POA is also a major distribution point for liquid fuels. On average, two-thirds of the fuel for air carriers at Ted Stevens Anchorage International Airport, and two-thirds of the fuel used by the U.S. military and federal government agencies in Alaska are delivered through the Port. This includes 100% of the jet fuel for Joint Base Elmendorf-Richardson (UAA, 2011).

The primary types of marine transportation moving freight and passengers to Alaska include:

 Railcar barges operating from Tacoma, Washington, and Prince Rupert, British Columbia;  Ocean vessels providing roll-on/roll-off services for highway trailers operating from Tacoma, Washington;  Container vessels originating in Tacoma, Washington;  Ferries operating in Southeast, Southcentral, and Southwestern Alaska; and  Barges operating from the Pacific Northwest primarily to Southeast and Southcentral Alaska.

Key transportation providers serving this market are Totem Ocean Trailer Express (TOTE), Horizon Lines, CN AquaTrain, and Lynden Transportation. Railcar barge movements destined for the Port of Whittier, Alaska, connect with the Alaska Railroad for movement of goods to Anchorage, Fairbanks, and other inland destinations. A number of barge companies deliver goods to Southwest and Western Alaska from

the Port of Anchorage or from the Pacific Northwest. In addition to fuel delivered from refineries in Anacortes, Washington, some fuel is delivered from Asia.

The majority of this capacity serves Southcentral and Interior Alaska (and to a lesser extent Southeast Alaska), accessed primarily through the Ports of Anchorage and Whittier. The Port of Whittier, while served weekly by Alaska Marine Lines barge service, is predominantly used for delivery of railcar barges via the AquaTrain connecting with the Alaska Railroad. Based on the schedule and equipment of marine transportation to Southcentral and Southeast Alaska, the estimated total freight capacity of these service providers is approximately 4.7 million short tons annually (QGI, 2006). Estimates of imports and exports by water transportation are shown in Figure 3 and Figure 4. For more information on companies shipping goods to and from Alaska, see Appendix A.

Figure 3. Estimate of imports to Alaska via water transportation, thousands of short tons, 2010

Figure 4. Estimate of exports from Alaska via water transportation, thousands of short tons, 2010

Ferries The Alaska Marine Highway System (AMHS) operates 11 vessels serving 32 ports that transport more than 300,000 passengers, 100,000 cars, and 3,400 freight vehicles annually. The AMHS routes stretch over 3,700 miles serving Southeast Alaska, Prince William Sound, Kodiak Island, and the Aleutian Islands (Figure 5). The AMHS plays an important role in the economies of these regions and in Alaska’s transportation system (Metz et al., 2011).

The Inter-Island Ferry Authority (IFA) was formed in 1997 to improve transportation to island communities in southern Southeast Alaska. The Prince of Wales Island communities of Craig, Klawock, Thorne Bay, and Coffman Cove joined in a coalition with Wrangell and Petersburg to create the IFA; Hydaburg joined the group in 2010. The IFA is a public corporation organized under Alaska's Municipal Port Authority Act and governed by a Board of Directors.

The IFA development plan includes both the Hollis-Ketchikan and Coffman Cove-Wrangell-Petersburg passenger/vehicle ferry routes. Alaska Department of Transportation and Public Facilities (ADOT&PF) support for both routes was received in 1998. Alaska's congressional delegation secured funding for the

two planned IFA vessels. The M/V Prince of Wales inaugurated daily scheduled round-trip service between Hollis and Ketchikan in January 2002 (Figure 6). A sister vessel, the M/V Stikine, provided round-trip service from Coffman Cove to Wrangell and Petersburg for three summers (2006, 2007, and 2008), but this service is now on hold. The IFA ferries currently connect with vessels of the AMHS at Ketchikan.

Figure 5. Alaska Marine Highway System communities and routes Source: ADOT&PF, AMHS, 2012

Figure 6. Inter-Island Ferry Authority communities and routes Source: Inter-Island Ferry Authority, 2012

Trucking Trucking’s share of transportation employment in Alaska is considerably smaller than it is elsewhere in the country. Nationwide, the trucking industry employs over a third of all transportation workers, compared with about 15% in Alaska. But the rest of the nation enjoys a vast network of interstate and secondary highways that connect most communities to the road system (Fried and Keith, 2005; U.S. DOT, 2010). Alaska is connected to the rest of the nation via the Alaska Highway, but does not have the well-developed road system of states in the Lower 48 (Figure 7). As a result, transportation by truck is a smaller portion of the transportation industry in Alaska than it is nationally (Fried and Keith, 2005). In this analysis we estimate freight movement and fuel use for Alaska long-haul trucking only; we do not estimate fuel use by local delivery trucks transporting freight. The companies that conduct these operations and the segments traveled are shown in Table 1.

Figure 7. Map of Alaska major roadways Source: ADOT&PF, 2012

Table 1. Long-haul trucking companies providing services in Alaska Anchorage to: Fairbanks to: Alaska to: Company Seward Soldotna/Homer Valdez Fairbanks SE AK Prudhoe Bay Lower 48/Canada AirLand X X X X American Fast Freight X X X Bob Benson X X Carlile X X X X X City Express X X X Husky Haulers X Lynden X X X Midnight Sun X X X Pacific Alaska Freightways X X X Sourdough X X Weaver Brothers X X X Wilson Brothers X Source: Company websites

Railroads Two railroads serve Alaska. One is the publically owned Alaska Railroad, and the other is the privately owned White Pass and Yukon Route Railroad. The Alaska Railroad is an independent corporation serving ports and communities from the Gulf of Alaska to Fairbanks (Figure 8).

The State of Alaska bought the railroad from the federal government in 1985. The Alaska Railroad is governed by a seven-member board of directors appointed by the governor of Alaska, and is mandated to be self-sustaining and responsible for all its financial and legal obligations (ADOT&PF, 2008). Alaska has 632 total railway miles—611 public miles owned by the Alaska Railroad Corporation and about 21 miles privately owned by the White Pass and Yukon Route Railroad, providing links into Canada.

The Alaska Railroad is a major part of the transportation network, both within the state and between Alaska and the Lower 48. It connects with rail service from the rest of the U. S. and Canada via its barge facilities in Whittier, and ships coal and naphtha to Asia via the Port of Seward. The railroad carries both passengers and freight, but large volumes of a variety of freight account for most of its operating revenue (Verrelli, 2012). In recent years, petroleum products hauled from the North Pole refinery to the Anchorage area have made up much of the railroad’s freight revenue. The Alaska Railroad carries several hundred thousand tons of coal per year between Healy and Seward for overseas export to Asia and South America; it hauls coal from Healy to Fairbanks and a significant portion of the gravel used in the Anchorage bowl from the Matanuska-Susitna Valley. The railroad can carry these large volumes of freight more efficiently and at lower cost than trucks can (ICF International, 2009; Tuck and Killorin, 2004).

The Alaska Railroad provides passenger service to tourists during the summer season. The railroad is part of the tourist infrastructure, providing access to Denali National Park and other destinations.

The White Pass and Yukon Route Railroad is a narrow-gauge railroad that operates solely for tourism, between Skagway, Alaska, and Carcross, Yukon, each year from May to September. A wholly owned subsidiary of Tri-White Corporation based in Toronto, Ontario, the White Pass and Yukon Route generated $18.2 million in 2006, with 431,249 passenger trips (ADOT&PF, 2008). Though this railroad was originally developed to serve Yukon gold mining, and served as an ore-carrying railroad as recently as the 1970s, the owners recently expressed limited interest in resuming the railroad‘s ore-carrying capacity (ADOT&PF, 2008).

Figure 8. Alaska Railroad route map Source: Alaska Railroad, 2012. The red line is the Alaska Railroad between Fairbanks and Seward.

Aviation Airports, seaplane bases, and heliports located in remote geographic regions of Alaska are critical to the movement of passengers and freight within the state and to and from other national and international destinations. The Alaska Aviation System Plan (ADOT&PF, 2011) included an assessment of the contribution of the aviation industry to the Alaska economy (Northern Economics, Inc., 2009, 2011). The aviation industry, as defined in the statewide analysis, includes all the businesses and organizations located at an airport. Spending by these on-site entities supports local businesses and employs Alaskans for its year-round operations, contributing $3.5 billion directly and indirectly to the state’s economy. This dollar amount equals approximately 8% of the state’s $42 billion 2007 gross state product (GSP), a 40% higher share than the aviation industry’s contribution to the U.S. economy. The analysis also estimated that the aviation industry creates more than 27,000 on-site jobs and almost 20,000 off-site jobs, which represents about 10% of jobs in Alaska—again over 40% more than the national percentage of jobs in aviation.

The primary reasons for the prominence of the aviation industry in the Alaska economy are the state’s large geographic area, remoteness, and lack of connected roads. According to the 2011 Alaska Aviation System Plan, 82% of the communities in Alaska are not connected to a highway or road system and rely on air service to transport goods and passengers. As a result, the state has a large aviation network, with 10,000 pilots operating in 700 registered airports and 1,200 airstrips across more than three million square miles (Alaska Department of Labor and Workforce Development, Research and Analysis, 2012).

According to a 2009 economic study (Northern Economics, Inc., 2009), the average number of annual enplanements per capita for off-road communities in Alaska is 14.6, eight times higher than the number of annual enplanements per capita for even the next highest state—Idaho at 1.8—and more than 30 times higher than the lowest comparison group—Montana at 0.5 enplanements per person per year. The number of freight pounds per capita for Alaska is 39 times higher than that of rural communities in the next-highest surveyed state. Alaska communities in the study averaged 1,096 pounds of airfreight per capita in 2007, while rural communities in Oregon averaged 28 pounds. Rural communities in Montana averaged just 2 pounds of airfreight per person in 2007. Alaska, and especially remote rural communities not connected to roads, clearly depends on air to transport passengers and goods.

These transportation services are provided by 271 commercial operators in Alaska and over 10,000 licensed pilots, of which more than 2,800 are commercial pilots. Commercial carriers fly over 835,000 hours annually, including 420,000 scheduled flight hours and 415,000 unscheduled flight hours (Alaska Air Carriers Association, 2012). Alaska has a fleet of 10,947 aircraft, of which 40% are based in Anchorage and 85% are single-engine fixed wing. Anchorage records 1.6 million landings annually, including 2.8 million metric tons of cargo, making it the third highest among world airports in cargo volume (Alaska Air Carriers Association, 2012).

Research Approach This analysis necessitated the collection of a considerable amount of proprietary data from transportation companies. To the extent possible, we collected information for the years 2006 through 2010 directly from marine shipping, barge, and trucking companies. We obtained aviation data from the U.S. Department of Transportation, Research and Innovative Technology Administration (RITA), Bureau of Transportation Statistics (BTS). We downloaded aviation data from the RITA website and used that data to estimate fuel consumption and costs by aviation fleet type (U.S. DOT, RITA, BTS, 2010).

The statistics sections of the Alaska Marine Highway System (AMHS) and the Alaska Railroad Corporation provided us with data. The Alaska Inter-Island Ferry Authority (IFA) also provided requested data. The IFA and AMHS data were the most complete data we received.

For barges and trucking, only one company in each subsector provided data. We used these data in conjunction with secondary data to model the barge and trucking subsectors.

A recent National Cooperative Freight Research Program (NHCRP) report (Holguín-Veras, José, et al, 2013, found that there is a lack of freight cost data for the various modes of freight transportation, and that no single source can provide the key cost data for any mode. In most cases, some data are available

in the reports published by public-sector agencies, trade groups, and research universities. However, because these data were collected in response to the needs of specific projects, they cannot replace data formally collected as part of regularly scheduled data collection efforts. Publicly available cost data for air freight and terminals are practically nonexistent. Unfortunately, no single data source could fill all the gaps in freight cost data. Our data collection efforts and analyses were hampered by this data issue.

For each transportation mode, we attempted to estimate fuel use and cost as a total and per ton-mile and/or passenger-mile, as applicable. A ton-mile is defined as one ton (2,000 pounds) transported one statute mile. Ton-miles are computed by multiplying the net weight of the carried freight times the segment mileage for a shipment. For example, if the Alaska Railroad carries 26,280 passengers on the 112-mile Anchorage-to-Seward rail segment and uses 81,715 gallons of fuel, then the passenger-miles per gallon of fuel is 36:

(26,280 * 112)/81,715 = 36 passenger-miles per gallon

Similarly, if the railroad moves 2,361,900 tons of gravel on the 55-mile Wasilla-to-Anchorage rail segment, using 199,164 gallons of fuel (because the cars travel empty one way), then the ton-miles per gallon of fuel is 652:

(2,361,900 * 55)/ 199,164 = 652 ton-miles per gallon of fuel

We did not include the weight of the ship, plane, train, or truck in making these estimates. However, the weight difference and fuel use intensity are reflected in the average fuel-use-per-mile statistics that we calculate. Fuel and energy use while in ports, airports, rail yards, and truck depots are not included in this analysis. Fuel use is for the transportation of freight and passengers. For each transportation mode, we also calculated total CO2 emissions and emissions per ton-mile and/or passenger-mile (U.S. DOE, EIA, 2012b).

We converted all fuel prices to 2011 dollars using the U.S. Consumer Price Index (CPI) as reported by the Alaska Department of Labor and Workforce Development (ADLWD, 2012).

Details on each transportation sector’s data and modeling are presented in the following sections. Because of the considerable differences in the data provided and the subsequent need to construct models to develop final datasets, the methods sections differ considerably in length and detail. The precision of the data also varies considerably, and readers should note the limitations of the data when using the results.

Water Transportation Marine Ships Marine shipping companies provided monthly data on northbound and southbound tonnage and gallons of fuel used during 2006 through 2010. One company provided monthly fuel prices per barrel, while the other provided the total annual cost of fuel. These data were used to estimate marine shipping fuel prices and to calculate ton-miles of freight moved per gallon of fuel and the cost of fuel per ton-mile. To

avoid potential release of proprietary data, we present marine shipping results aggregated or as part of other water transportation statistics.

Barges In contrast with other transportation modes where we received considerable information from a number of companies—or the data were publically available through government reporting requirements—only one barge company provided an aggregation of monthly data for the movement of freight. We used this barge company data as a prototype to construct a barge fuel-use model. We estimated fuel use for regional barge shipments by taking the number of additional barge trips by travel segments from the U.S. Army Corps of Engineers’ published Waterborne Statistics of freight movement by port (U.S. Army Corps of Engineers, 2012a, b) and published freight schedules of barge companies serving Alaska. We used the prototype barge company’s information in conjunction with U.S. Army Corps of Engineers data to estimate types of tugs and barges used and their fuel consumption per mile traveled. We made these estimates in the absence of publically available or proprietary data, but we believe they are reasonable, carefully developed estimates. Still, they are merely estimates. Details on these calculations are provided in Appendix B.

Ferries We used annual reports of the Alaska Marine Highway System (AMHS) to estimate monthly data on northbound and southbound passengers and freight in between ports. The Inter-Island Ferry Authority also provided monthly data for 2006 through 2010. We estimated short tons of freight based on the average weight per vehicle class and the number of vehicles reported in different vehicle classes in the AMHS annual reports. Fuel cost information came directly from fuel purchase invoices. Fuel invoice information included the date and location of the fuel purchase and the receiving vessel. We allocated fuel consumption between ports based on the mileage between ports of a particular vessel port of call. Finally, we were able to estimate monthly fuel consumption and fuel cost in between ports, which was matched to the monthly vessel port-to-port data. We have used the resulting dataset to calculate ton- miles of freight moved per gallon of fuel and the cost of fuel per ton-mile for shipping by ferry, as well as passenger-mile per gallon of fuel.

Recognizing the differences in ferry configuration and fuel usage when allocating fuel to freight and passengers, we separated the fleet into four groups: (1) high-speed catamarans, (2) Aleutian chain, (3) southern Southeast day boats (IFA and Lituya), and (4) the remaining mainline ferries. Our consultation with a number of marine architects and review of the literature indicated little agreement on how to allocate fuel used to carry freight and passengers on mixed-use ferries. Based on these discussions and the literature review, we settled on 90% fuel allocated to passengers and 10% fuel allocated to freight on catamarans and day boats, and a 50%-50% split between freight and passengers on main line and Aleutian chain ferries.

Land Transportation Railroad Though the Alaska Railroad provided data on tonnage, passengers, and gallons of fuel used by departure and destination for 2006 through 2010, it did not provide fuel cost information. As a result, we substituted Anchorage refinery fuel prices reported by the Oil Price Information Service (OPIS, multiple years) for the missing fuel price information. We also did not receive fuel cost/price information from Alaska trucking companies, so we used the same OPIS prices for a substitute. Thus, while our estimates of fuel prices per gallon are not accurate, they are comparable for rail and truck, which are the two primary competitors for land shipping in the Alaska Railbelt. Our results compare the relative efficiency of freight movement by the two modes, rather than the definitive cost over the period of analysis (GAO, 2011; Center for Neighborhood Technology, 2013).

Trucking As was true of barge companies, only one long-haul trucking company shared data on tonnage and fuel use by destination. We estimated market share and expanded the data for an all-Alaska tonnage and fuel use evaluation. We used those data to calculate ton-miles of freight moved per gallon of fuel and the cost of fuel per ton-mile for shipping by truck.

Aviation To estimate fuel used in aviation, we initially attempted to use the U.S. Department of Transportation, Bureau of Transportation Statistics (BTS) data, available for download at www.transtats.bts.gov. Our initial analysis was based on three BTS data sources: T100 segment data, Schedule T2, and schedule P- 12(a) (U.S. DOT, RITA, BTS, 2012 a, b, c and d). T100 segment data show the monthly number of flights for a city-pair, also called a flight segment. The T100 data are not based on a sample or survey; they represent a 100% census. All carriers except those with $20 million or less in annual operating revenue submit quarterly balance sheets and fuel reports. Schedule T2 provides quarterly air carrier traffic and capacity statistics by aircraft type and carrier, including the amount of aircraft fuel issued (not used). Schedule P-12(a) shows monthly fuel consumption and fuel cost by carrier, but does not disaggregate fuel consumption and cost by each carrier’s aircraft types.

First, we combined ten years of T100 segment data from 2000 to 2010 for flights originating in Alaska, thus including flights within Alaska and the first segment of flights originating in Alaska for out-of-state destinations. The data included both scheduled and unscheduled flights. We then followed the same procedure to combine the same ten years for Schedule T2 and Schedule P-12(a).

In an effort to link the T100, T2, and P-12(a), we tried to estimate fuel used per T100 segment. Therefore, we first used the T2 data to calculate carrier and aircraft-group-specific fuel efficiencies per mile and per hour of flight for each quarter of the year between 2000 and 2010. To do so, we applied the mean quarterly gallons per mile flown or hour flown for each carrier by aircraft type. For cases with missing data on the amount of fuel issued, we applied different approaches, depending on whether we knew the aircraft type and/or aircraft group.

We calculated fuel cost per segment based on Schedule P-12(a). We divided the monthly carrier-specific fuel cost by the monthly carrier-specific fuel amount in gallons, which equals the nationwide monthly average fuel price per carrier. We then applied this carrier-specific monthly fuel price to the fuel consumption per segment to arrive at the fuel cost per segment.

We tested our model results by comparing them with the U.S. Department of Energy (DOE), Energy Information Administration (EIA), State Energy Data System (SEDS) estimate for Alaska aviation fuel use. This comparison indicated that our model estimates using publically available, unlinked data were about an order of magnitude larger. Others have identified similar difficulties using BTS data as well as the non-existence of air freight cost data (Holguín-Veras, José, et al., 2013; Peeters et al., 2005; Siebe, 2012; Lee et al., 2001). Peeters et al. (2005) argues that fuel consumption data from BTS do not take into account the fact that planes load additional reserve fuel, which is issued but not used. Lee et al. (2001) mentioned this as well. Thus, the variable AIRCRAFT_FUELS_921 represents the gallons of fuel issued but not necessarily used. Peeters et al. (2005) write that in order to use the BTS Schedule T-2 data alone, one has to correct for fuel reserves. The authors note that piston-engine aircrafts carry about three hours of reserve fuel, while jets take extra fuel for about 200 nautical miles (230 miles). These factors exacerbate the problems of matching the unlinked fuel used (T-2), segments flown (T100), and fuel cost (P-12[a]) datasets. In Alaska, the problem of discrepancy between the reported fuel issued and fuel used is made worse by the fact that aircraft almost exclusively fuel in the large airports of Anchorage and Fairbanks and very rarely refuel at airports in rural Alaska, where fuel is often more than double the price (Cadavoa, 2010). Also, additional fuel beyond that used in flight and carried for reserve is often transported to rural fuel depots, where it is stored in the airline’s fuel cache for emergencies and other purposes, like heating airline-owned facilities.

Our fallback for estimating fuel use was to use data from the U.S. DOE EIA SEDS. With this data, we estimated total fuel aviation use, but we used BTS data to allocate the fuel to scheduled and unscheduled intra-state flights and flights exiting Alaska by carrier types—passenger, cargo, mixed passenger and cargo, and seaplanes. We used the estimate of fuel used by exiting flights to estimate fuel used by scheduled and unscheduled flights entering Alaska. The variables and data used from specific data sources are shown in Appendix C.

While this method provides a more reasonable estimate of fuel used by different carrier configurations, the level of aggregation is too great to estimate fuel use per ton-mile or per passenger-mile, which does not facilitate a comparison of energy efficiency or costs across transportation modes. To find potential proxies, we consulted the literature.

Economic Impact Analysis of Fuel Price Changes on the Alaska Economy The foundation of modern input-output analysis is based on work started in the 1930s by Wassily Leontief (Leontief, 1936, 1966). Economic theory abstractly describes the relationships between prices and quantities with respect to supply and demand in a market economy. The ways that these relationships unfold in reality, however, are based on innumerable individual transactions involving a vast array of inputs, products, and services. By collecting, aggregating, and tabulating detailed industrial

output data into a matrix, in which the output of every industry may serve as the input to a variety of other industries in an economy, Leontief created an analytic tool that bridges the gap between the abstraction of economic theory and the empirical detail found in economic data.

Table 2 provides an illustration of this input-output transactions tool. The columns represent the variety of industrial input requirements (demand), and the rows represent the distribution of industrial output (supply). In addition to the square industry-by-industry transaction matrix (producing sectors), the model includes a vector at the bottom for value added and a vector along the right-hand side of the matrix for final demand. The value-added vector comprises primary factor inputs to production, such as capital and labor services. The final demand vector comprises the components that make up gross domestic product (GDP): consumption, investment, imports, exports, and government. Because of the basic accounting premise that all outputs in an economy must equal all inputs, the total output for a given industry can be calculated as either the column sum of intermediate inputs and value added, or as the row sum of intermediate and final demand for its output. In addition, total value added (the row sum of the vector), which represents all the income in the economy, must equal total final demand (the column sum of the vector), which represents the output of the economy.

Table 2. Illustrative input-output transactions table (in millions of dollars) Producing Sector Consuming Sector Industry Industry Industry Exports Households Total Final Total Sales A B C Demand (A+B+C+TFD) Producing Sectors Industry A 10 5 3 1 12 13 31 Industry B 3 9 8 1 4 5 25 Industry C 8 4 6 3 3 6 24 Primary Inputs Value Added 10 7 7 0 8 8 32 Total Inputs 31 25 24 5 27 32 112

The values in the matrix and the vector represent dollar transaction values, each comprising a price component and a quantity component. The nominal transaction values in the matrix and the vector can be converted into coefficients by dividing each column value by the value of total industry output. The calculated coefficients represent the proportions of inputs required to produce a single unit of output, and each column sums to one. This matrix of coefficient values, known as the matrix or the direct requirements matrix, can be thought of as the production "recipes" for each industry. When viewed as a whole, the entire matrix provides a snapshot of the current technological state of an economy. For more details on input-output modeling and terminology, see Appendix D.

IMPLAN IMPLAN (IMpact analysis for PLANning) is a system for conducting economic analyses based on national input-output (I/O) structural matrices (MIG, Inc., 2011). IMPLAN was originally developed by the U.S.

Forest Service and has gained wide acceptance in a variety of impact assessment applications. In addition to the U.S. Forest Service, users of IMPLAN have included the U.S. Army Corps of Engineers, the National Park Service, the Soil Conservation Service, the Federal Emergency Management Agency, the Bureau of Land Management, universities, and numerous state and regional planning agencies.

The basic IMPLAN model performs an I/O analysis for a given region in terms of as many as 509 economic sectors (257 for Alaska), roughly corresponding to NAIC (North American Industry Classification) codes. In addition, IMPLAN allows the analyst to add custom sectors for a particular application. Impacts are specified in terms of output, income, and employment.

The economic impacts estimated by input-output models reflect the direct expenditures of a particular sector (study sector) and account for the “ripple effect” of economic activity resulting from that sector. Employees of the study sector and local businesses from which the study sector purchased goods and services continue to spend at least some percentage of these monies locally, spurring additional economic impacts. The initial expenditure essentially spurs a chain of indirect and induced spending. Input-output models use a series of “multipliers” to estimate the economic impacts associated with each initial dollar of direct spending. We use IMPLAN to analyze how the fuel price changes that occurred from 2009 to 2010 (approximately 28%) affected the Alaska economy as depicted in 2008 before the price increases. To the extent possible, we analyzed transportation fuel prices from 2006 to 2010, because that period witnessed dramatic changes in prices (Figure 9).

Figure 9. Annual crude oil prices, 2000–2012 Source: U.S. Energy Information Administration, Cushing, OK WTI Spot Price FOB (Dollars per Barrel).

Findings and Applications Fuel Use and Efficiency by Mode Water Transportation Marine ships and barges carry an estimated 146–186 ton-miles per gallon of fuel (a ton-mile is the movement of one ton of freight one mile); they are well laden despite the fact that they primarily carry

freight into Alaska and return considerably less laden. Despite already reflecting relatively higher fuel efficiency per ton-mile, the data indicate increasing efforts to move freight even more efficiently as fuel prices increased in 2008 (Winebrake, James J. and James J. Corbett, 2010) . However, because of the confidentiality of proprietary data, specific details cannot be presented. Table 3 and Table 4 show our estimates of average fuel use, costs, and fuel use per ton-mile of freight shipped for marine ships and barges.

Table 5 shows regional barge fuel use estimates from our barge fuel-use model.

Marine Ships

Table 3. Estimated fuel usage and costs for freight movement by marine vessels 2006–2010, 2011$

Average ship miles per gallon 0.02 Average ship ton-miles per gallon 146 Average fuel costs per mile $102 Average fuel costs per ton-mile $0.012 Source: OPIS Anacortes fuel prices, author calculations.

Barges

Table 4. Estimated tons, fuel usage and costs for freight movement by barge 2006–2012, 2011$

Average barge miles per gallon 0.04 Average barge ton-miles per gallon 186 Average fuel costs per mile $51 Average fuel costs per ton-mile $0.016 Source: Marine fuel prices, U.S. Waterborne statistics, author calculations.

Table 5. Estimated regional fuel usage for freight movement by barge 2006–2012, 2011$

Southcentral 2006 2007 2008 2009 2010 Fuel (gallons) 15,469,855 15,405,158 15,265,487 15,154,696 13,438,447 Fuel costs $45,288,048 $45,420,378 $61,122,310 $38,201,410 $39,984,049 Distance 843,271 839,744 832,131 826,092 732,538 Miles/gallon 0.05 0.05 0.05 0.05 0.05 Fuel costs/mile $54 $54 $73 $46 $55 Southeast Fuel (gallons) 9,269,933 9,570,533 8,568,850 9,844,300 9,194,950 Fuel costs $27,022,116 $28,254,127 $33,854,190 $24,918,379 $27,427,971 Distance 640,059 660,814 591,651 679,717 634,882 Miles/gallon 0.07 0.07 0.07 0.07 0.07 Fuel costs/mile $42 $43 $57 $37 $43 Western Cargo 2,272,073 2,386,748 1,977,094 2,037,358 2,106,688 Fuel (gallons) 9,311,500 9,294,750 9,623,375 9,226,000 8,878,000 Fuel costs $27,535,811 $26,928,952 $39,333,492 $22,615,153 $26,373,031 Distance 428,619 427,848 442,975 424,684 408,665 Miles/gallon 0.05 0.05 0.05 0.05 0.05 Fuel costs/mile $64 $63 $89 $53 $65 Inland* Fuel (gallons) 119,880 119,880 119,880 119,880 119,880 Fuel costs $346,837 $361,320 $477,498 $313,125 $369,396 Distance 13,796 13,796 13,796 13,796 13,796 Miles/gallon 0.12 0.12 0.12 0.12 0.12 Fuel costs/mile $25 $26 $35 $23 $27 Total Gallons 34,171,168 34,390,321 33,577,592 34,344,876 31,631,277 Cost (2011$) $100,192,812 $100,964,777 $134,787,490 $86,048,067 $94,154,448 * Inland waterways figures are minimum estimates based on 2010 U.S. Census population. Sources: U.S. Waterborne Statistics, multiple years; Fisheries Economics Data Program, Monthly Marine Fuel Prices, http://www.psmfc.org/efin/data/fuel.html, author calculations.

Ferries Ferries serve as part of the Alaska highway system in locations where no roads exist—primarily Southeast Alaska, Prince William Sound, and the Aleutian chain. They also connect Southeast and Southcentral Alaska to the Lower 48 highway system, with voyages to and from Bellingham, Washington, and Prince Rupert, British Columbia. Ferries are constructed to handle rough seas and have traditionally carried passenger vehicles and freight—and both those factors reduce their fuel-use efficiency. However, ferries have always been under pressure not to compete with the private sector for freight, so their tariffs are set to be non-competitive, and as a result, they primarily carry freight cargo to smaller communities where barge service is not economical. So it can be expected that the fuel and

operating efficiencies are lower for AMHS ferries than for private shipping and barge companies (Table 6 and

Table 7). Given the age and configurations of Alaska ferries, the policy of avoiding direct competition with the private sector, and the schedules operated to meet the needs of the traveling public, it is unlikely that ferries could ever be as fuel-efficient as their private-sector counterparts. The AMHS ferries operate on less than a tenth of the ton-miles per gallon as barges and ships, primarily because they use less of their capacity. Keep in mind that ferries provide essential transportation services to locations not connected to roads.

For the years analyzed, passenger-miles per gallon for ferries ranged from 11 to 13 (Table 6 and Table 8), making ferries the least fuel efficient for moving passengers of all the modes analyzed, including air. This lack of fuel efficiency results from low capacity factors and high fuel use. Compared with national figures, Alaska ferries on average have lower fuel efficiencies than passenger vehicles have (Ghanta, 2010). Day boat and catamaran service appear to improve these statistics, because more space is dedicated to passengers and the service schedule is more closely tied to travelers’ schedules. Table 9 provides more details on fuel use by the four groups of ferry service, as discussed in the methods section: (1) high-speed catamarans, (2) Aleutian chain, (3) southern Southeast day boats (IFA and Lituya), and (4) the remaining mainline ferries. Considerably more analysis is available for the ferry system because of the data received, and can be provided upon request.

Table 6. Passengers, freight, fuel usage, and costs for the Alaska Marine Highway System and Inter-Island Ferry Authority 2007–2010, 2011$ 2007 2008 2009 2010 Total passengers 589,599 603,890 563,302 579,508 Total cargo deck freight (short tons) 688,646 666,529 658,575 664,407 Fuel use, passenger share 7,089,193 6,320,160 5,912,334 6,153,616 Fuel use, freight share 4,857,297 4,135,185 4,237,976 4,296,995 Total fuel use (gallons) 11,946,490 10,455,345 10,150,310 10,450,611 Total fuel cost (2011$) $30,493,483 $38,107,148 $22,896,161 $28,345,120 Total miles 726,196 666,298 640,738 645,955 Miles/gallon 0.06 0.06 0.06 0.06 Fuel cost per mile $42 $57 $36 $44 Ton miles/gallon 18 18 20 20 Fuel cost per ton-mile $2.28 $3.28 $1.74 $2.18 Ton miles (freight) 88,844,314 78,258,524 77,939,792 80,028,674 Passenger miles 68,401,657 65,009,076 61,092,295 64,055,703 Passenger miles/ gallon 11 11 12 13 Fuel cost per passenger-mile $3.81 $5.25 $2.81 $3.45 Passenger % utilization (AMHS) 24% 27% 28% 28% Sources: ADOT&PF, AMHS; IFA; U.S. DOE, EIA; author calculations.

Table 7. Estimated tons, fuel usage, and costs for freight movement for the Alaska Marine Highway System and Inter-Island Ferry Authority 2007–2010, 2011$

Average miles per gallon 0.06 Average ton-miles per gallon 19 Average fuel costs per mile of ship operation $45 Average fuel costs per ton-mile $2.37 Sources: ADOT&PF, AMHS; IFA; U.S. DOE, EIA; author calculations.

Table 8. Estimated tons, fuel usage, and costs for passenger movement by the Alaska Marine Highway System and Inter-Island Ferry Authority, 2007–2010, 2011$

Average miles per gallon 0.06 Average passengers-miles per gallon 12 Average fuel costs per mile of ship operation $45 Average fuel costs per passengers-mile $3.83 Sources: ADOT&PF, AMHS; IFA; U.S. DOE, EIA; author calculations.

Table 9. Annual passengers and freight by segment type for the Alaska Marine Highway System and Inter-Island Ferry Authority segment types 2007–2010, 2011$

Ferry "type" 2007 2008 2009 2010 Total Passengers (Count) Aleutian 35,096 33,129 34,182 35,581 Catamaran 88,880 74,997 55,490 62,299 IIF 88,620 85,665 81,878 79,800 Main Line 377,003 410,099 391,752 401,828 Total 589,599 603,890 563,302 579,508 Total Freight (Short Tons) Aleutian 43,400 42,843 48,014 45,802 Catamaran 84,287 70,791 55,403 62,095 IIF 81,332 68,838 66,498 65,185 Main Line 479,627 484,057 488,660 491,325 Total 688,646 666,529 658,575 664,407 Gallons per Segment Aleutian 1,414,387 988,542 1,146,682 1,154,950 Catamaran 2,283,737 2,324,169 1,697,233 1,989,729 IIF 506,132 407,050 395,715 331,048 Main Line 7,742,234 6,735,584 6,910,680 6,974,884 Total 11,946,490 10,455,345 10,150,310 10,450,611 Fuel Cost (2011$) Aleutian $3,625,386 $3,635,715 $2,842,751 $3,377,013 Catamaran $5,955,128 $9,338,305 $3,831,519 $5,500,718 IIF $1,352,102 $1,379,505 $798,404 $862,022 Main Line $19,560,866 $23,753,623 $15,423,486 $18,605,367 Total $30,493,483 $38,107,148 $22,896,161 $28,345,120 Statute Miles per Segment Aleutian 93,534 69,679 75,069 76,152 Catamaran 112,865 117,281 85,799 93,996 IIF 58,078 49,800 49,806 45,761 Main Line 461,719 429,538 430,064 430,046 Total 726,196 666,298 640,738 645,955 Sources: ADOT&PF, AMHS; IFA; U.S. DOE, EIA; author calculations.

Land Transportation

Railroad The Alaska Railroad provides freight and passenger service along the corridor from Fairbanks to Seward and Whittier. Approximately a third of the tonnage of the railroad’s freight service was refined petroleum products in 2007. However, with production declines at the Flint Hills refinery, petroleum products’ share of total freight has declined. The railroad showed ton-mile per gallon freight efficiencies from 256 to 311 tons per mile during the study period (Table 10, Table 12, and Figure 10). The national rail service company, CSX, advertises almost 500 ton-miles per gallon, but Lower 48 trains and segment distances are considerably longer (CSX, 2012). Also, it is not clear whether CSX calculations include only

full loads and exclude empty back hauls; empty back hauls reduce the ton-miles per gallon for the Alaska Railroad.

Table 10. Fuel usage and costs for freight movement by the Alaska Railroad 2007–2010, 2011$

Fuel Ton-mile/ Fuel cost/

Year Short tons Gallons Cost gallon ton-mile 2007 6,592,506 4,342,932 $11,135,924 260 $0.010 2008 6,897,737 3,241,907 $11,083,097 288 $0.012 2009 6,626,029 3,618,137 $8,341,123 311 $0.007 2010 7,069,781 3,993,404 $11,042,024 256 $0.011 Average 6,796,513 3,799,095 $10,400,542 279 $0.010 Source: Alaska Railroad data, OPIS Anchorage fuel prices, author calculations.

Table 11. Alaska Railroad freight transport in short tons, petroleum and all other goods Short tons Petroleum AOG Total 2007 2,202,162 4,390,344 6,592,506 2008 1,911,157 4,986,580 6,897,737 2009 1,657,763 5,179,170 6,836,933 2010 1,253,894 5,815,887 7,069,781 Source: Alaska Railroad data, multiple years.

8,000,000

7,000,000

6,000,000

5,000,000

4,000,000

3,000,000

2,000,000

1,000,000

- 2007 2008 2009 2010 Petroleum AOG

Figure 10. Alaska Railroad freight transport in short tons, petroleum and all other goods (AOG), 2007–2009 Source: Alaska Railroad data

The estimates of fuel costs per mile and fuel costs per ton-mile (Table 12) are based on OPIS Anchorage refinery prices, because the Alaska Railroad provided no fuel price information. Most likely the OPIS prices are lower than the railroad’s actual fuel costs, but because it transports refined fuels for refineries and uses large quantities, the railroad may negotiate prices that are relatively close to OPIS wholesale prices. Also, in the absence of actual data, the OPIS prices are the most appropriate substitute. But the fuel costs provided in Table 12 might not be accurate, and should not be compared with other modes, like ferries, that provided actual price information.

Table 12. Average fuel usage and costs for freight movement by the Alaska Railroad 2007–2010, 2011$

Average miles per gallon 0.13 Average rail ton-miles per gallon 279 Average fuel costs per mile $21 Average fuel costs per ton-mile $0.01 Source: Alaska Railroad data, OPIS Anchorage fuel prices, author calculations.

The Alaska Railroad averages approximately 100 to 150 passenger-miles per gallon of fuel (Table 13 and Table 14). The low end of the range is approximately 40% higher than Amtrak’s national average, and the upper end is almost double (Ghanta, 2010). The pull contracts for the cruise ship companies taking visitors to Denali National Park and Preserve provide good utilization rates, as these are relatively long, well-used passenger trains.

Table 13. Fuel usage and costs for passenger movement by the Alaska Railroad 2007–2010, 2011$

Fuel Passenger-mile/ Fuel cost/ Year Passengers Gallons Cost gallon Passenger-mile 2007 670,868 1,220,758 $3,092,549 102 $0.02 2008 675,626 1,228,142 $4,879,996 123 $0.03 2009 586,149 1,195,411 $2,811,418 148 $0.01 2010 516,480 1,149,241 $3,158,125 107 $0.03 Average 612,281 1,198,388 $3,485,522 120 $0.02 Source: Alaska Railroad data, OPIS Anchorage fuel prices, author calculations.

Table 14. Average fuel usage and costs for passenger movement by the Alaska Railroad 2007–2010, 2011$

Average miles per gallon 0.2 Average rail passengers-miles per gallon 120 Average fuel costs per mile $15 Average fuel costs per passengers-mile $0.02 Source: Alaska Railroad data, OPIS Anchorage fuel prices, author calculations.

Trucks Our truck fuel-use estimates are based on limited information and should be used with caution. The fuel use per mile reported in Table 15 is higher than the national average of 6 miles per gallon and the average of 59 ton-miles per gallon (U.S. Department of Transportation, 2010). These higher Alaska fuel use rates most likely can be attributed to Alaska’s colder conditions and mountainous terrain, and to road systems that are not comparable to Lower 48 interstate highways.

Table 15. Estimated fuel usage and costs by long-haul trucks, 2011$

Average truck miles per gallon 4.5 Average truck ton-miles per gallon 48 Average fuel costs per mile $0.61 Average fuel costs per ton-mile $0.06 Source: OPIS Anchorage fuel prices, author calculations.

As was true for the Alaska Railroad, we did not receive any fuel cost information for trucks, so we substituted OPIS Anchorage refinery prices for truck diesel prices. Despite the fact that some companies haul fuel for refineries, the numerous trucking companies most likely purchase fuel in smaller quantities than the railroad does, so this fuel price estimate is more likely to be low for trucks than for the railroad. As a result, truck fuel costs should not be compared with those of other modes and compared cautiously with those of the railroad.

Aviation Table 16 provides information on estimated fuel used for scheduled and unscheduled intra-state flights, scheduled and non-scheduled flights exiting Alaska, and scheduled and unscheduled flights entering Alaska for the years 2005 through 2010 (additional details are available in Appendix D). The information includes the total reported fuel consumed, total fuel costs, estimated price of fuel per gallon, number of passengers, tons of freight and mail, and fuel used for the four primary carrier types providing service to Alaska markets and freight carriers traveling through Alaska.

The total fuel used by the aviation sector ranged from a low of approximately 1.6 billion gallons in 2009 during the recession to 2.4 billion gallons in 2007 prior to the recession and the dramatic increase in crude oil prices. Of this total, in 2010, scheduled intra-state flights used approximately 1.5%, nonscheduled intra-state flights used 0.4%, flights exiting Alaska used an estimated 48.8%, and flights entering Alaska used 49.3%. Cargo flights entering and exiting Alaska used about 10 times the amount of fuel as passenger flights entering and exiting Alaska.

As discussed in the methods section, the distance information is the summation of segments traveled, but does not account for the number of times each segment was traveled. As a result, it is not possible to calculate passenger-miles or ton-miles per gallon of fuel or fuel costs per mile, passenger-mile, or ton- mile for aviation.

Airlines spend 10% or more of their operating budgets on fuel purchases, making fuel the air industry’s second largest expense category, after labor cost (Wells, 2011; Ghanta, 2010). A literature review on fuel consumption by air carriers shows increasing fuel conservation efforts among air carriers. In the 20- year period from 1970 to 1990, fuel efficiency nearly doubled, increasing from 26.2 seat-miles per gallon (SMPG) in 1970 to 49 SMPG in 1989. In the 1990s, fuel efficiency reached 65–80 SMPG. Approaches to minimize fuel consumption include continuous descents from altitude, increased monitoring of fuel use, and improved effectiveness of aircraft loading (Peeters et al., 2005; Stolzer, 2002; Wells, 2011; Ghanta, 2010; Lee et al., 2001). Based on this review, we used 50 passenger-miles per gallon of fuel for comparison with other modes.

Table 16. Fuel usage and costs for intra-state scheduled air transportation in Alaska 2005–2010, 2011$ 2007 2008 2009 2010 Intra-state scheduled Total gallons 30,684,229 31,457,698 27,330,122 27,846,200 Totalcost [2011$] $79,832,197 $103,765,976 $55,388,981 $69,842,835 Average price [$/gal] $2.60 $3.30 $2.03 $2.51 Passengers 3,165,770 3,175,766 2,801,996 2,958,337 Freight 99,460 97,430 90,176 94,379 Mail 114,531 114,771 106,317 106,877

CO2 emissions [mt] 293,587 300,987 261,495 266,432 Intra-state unscheduled Total gallons 5,963,125 9,301,368 6,732,250 7,859,375 Totalcost [2011$] 17,190,711 32,761,806 16,001,674 17,258,454 Average price [$/gal] $2.88 $3.52 $2.38 $2.20 Passengers 195,941 185,469 197,444 170,207 Freight 19,601 21,541 35,116 19,025 Mail 565 373 151 112

CO2 emissions [mt] 57,055 88,995 64,414 75,199 Exiting (Scheduled & unscheduled) Total gallons 1,193,994,646 967,954,934 762,383,628 925,632,425 Totalcost [2011$] 2,856,751,124 3,063,036,030 1,427,105,264 2,169,344,617 Average price [$/gal] $2.39 $3.16 $1.87 $2.34 Passengers 2,215,930 2,062,806 1,844,701 1,893,169 Freight 3,005,120 2,456,141 1,988,147 2,472,836 Mail 4,778 4,196 2,664 4,009

CO2 emissions [mt] 11,411,037 9,250,826 7,283,021 8,847,839 Entering (Scheduled & unscheduled) Total gallons 1,207,421,777 982,127,718 772,386,783 934,194,928 Totalcost [2011$] $2,888,876,874 $3,107,884,964 $1,445,830,162 $2,189,412,000 Average price [$/gal] $2.39 $3.16 $1.87 $2.34 Passengers 2,215,930 2,062,806 1,844,701 1,893,169 Freight 3,005,120 2,456,141 1,988,147 2,472,836 Mail 4,778 4,196 2,664 4,009 CO2 emissions [mt] 11,539,361 9,386,276 7,378,581 8,929,685 Notes: Numbers in black are directly from the BTS and EIA data. Blue numbers were calculated from BTS and EIA data. Sources: U.S. DOT, BTS; U.S. DOE, EIA; FAA; author calculations.

Intermodal Fuel Use Comparisons As previously stated, the variability of available data makes direct comparisons of energy use difficult, but we can use our information to gain a broad understanding of the relative differences in fuel use across the modes to help illustrate differences in vulnerabilities to fuel price changes and potential emissions taxes. Despite the limitation of the data, Figure 11,Figure 12, and Figure 13 and Table 17 provide “overview” comparisons of fuel used and the estimated costs of fuel for the four travel modes analyzed from 2007 to 2010.

Table 18 through Table 20 and Figure 14 and Figure 15 compare estimated direct fuel use and costs across the four modes analyzed. Again, because of lack of precision, comparisons should be used in the context of general “ranges of magnitude.” This is especially true for rail and trucking, where we received no fuel price information. In the absence of this information, we substituted OPIS Anchorage refinery diesel prices (as discussed earlier). So these comparisons are imprecise for rail and trucking to the extent to which the two pay different mark-ups from wholesale refinery prices.

2,750

2,250

1,750

1,250

750 Millions

250

(250) 2007 2008 2009 2010 Air Water Trucks Rail (freight) Rail (passenger)

Figure 11. Comparison of annual fuel use, 2007–2010 Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations. The 2010 truck data are used in all four years in place of missing data.

$7,000

$6,000

$5,000

$4,000

$3,000 Millions, 2011$ Millions, $2,000

$1,000

$- 2007 2008 2009 2010 Air Water Trucks Rail (freight) Rail (passenger)

Figure 12. Comparison of annual fuel costs, 2007–2010, 2011$ Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations. The 2010 truck data are used in all four years in place of missing data.

$4.00

$3.50

Air (Intra, scheduled) $3.00 Air (Intra, unscheduled) Air (exiting) Water $2.50 Trucks Rail

$2.00

$1.50 2006 2007 2008 2009 2010

Figure 13. Comparison of annual fuel prices per gallon, 2007–2010, 2011$ Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations. The 2010 truck data are used in all four years in place of missing data.

Table 17. Estimated transportation fuel use and costs, Alaska 2007–2010

Aviation Scheduled intra-state 2007 2008 2009 2010 gallons 30,684,229 31,457,698 27,330,122 27,846,200 cost (2011$) $79,832,197 $103,765,976 $55,388,981 $69,842,835 Reported average price [$/gal] $2.60 $3.30 $2.03 $2.51 Non-scheduled intra-state gallons 5,963,125 9,301,368 6,732,250 7,859,375 cost (2011$) $17,190,711 $32,761,806 $16,001,674 $17,258,454 Reported average price [$/gal] $2.88 $3.52 $2.38 $2.20 Exiting (scheduled & non-scheduled) gallons 1,193,994,646 967,954,934 762,383,628 925,632,425 cost (2011$) $2,856,751,124 $3,063,036,030 $1,427,105,264 $2,169,344,617 Reported average price [$/gal] $2.39 $3.16 $1.87 $2.34 Entering (scheduled & non-scheduled) 1,207,421,777 982,127,718 772,386,783 934,194,928 $2,888,876,874 $3,107,884,964 $1,445,830,162 $2,189,412,000 $2.39 $3.16 $1.87 $2.34 Aviation totalsgallons 2,438,063,777 1,990,841,718 1,568,832,783 1,895,532,928 cost (2011$) $ 5,842,650,906 $ 6,307,448,776 $ 2,944,326,080 $ 4,445,857,906 Estimared average price [$/gal] $2.40 $3.17 $1.88 $2.35 Water Barges/Ships gallons 95,383,724 94,056,619 95,190,267 91,307,812 cost (2011$) $235,744,675 $309,973,422 $199,572,877 $227,870,320 gal. (intra-state & exiting) 78,243,753 77,260,621 78,012,015 73,839,840 Ferries gal. (intra-state & exiting) 11,946,490 10,455,345 10,150,310 10,450,611 cost (2011$) $30,493,483 $38,107,148 $22,896,161 $28,345,120 Total Water gallons 107,330,214 104,511,963 105,340,577 101,758,423 cost (2011$) $266,238,157 $348,080,571 $222,469,038 $256,215,440 Estimared average price [$/gal] $2.48 $3.33 $2.11 $2.52 Trucks (only one year of data) gallons 9,786,844 cost (2011$) $27,068,061 gal. (intra-state & exiting) 8,805,455 Estimared average price [$/gal] $2.77 Railroad gallons (freight) 4,342,932 3,241,907 3,618,137 3,993,404 cost (2011$) $10,291,981 $10,615,993 $7,959,087 $10,699,636 gallons (passengers) 1,220,758 1,228,142 1,195,411 1,149,241 Total rail cost (2011$) $2,858,178 $4,674,326 $2,682,651 $3,060,199 gallons 5,563,690 4,470,049 4,813,548 5,142,645 cost (2011$) $13,150,160 $15,290,319 $10,641,738 $13,759,835 Estimared average price [$/gal] $2.36 $3.42 $2.21 $2.68 Notes: Numbers in black are directly from the BTS and EIA data. Blue numbers were calculated from BTS and EIA data. Total weight includes freight, mail, and passengers. Aircraft type numbers are T100 data codes. Fuel prices in blue are estimates based on OPIS price data and limited proprietary data provided. Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

Table 18. Comparison of fuel use and costs per ton-mile for Alaska water transportation, 2007–2010, 2011$

Water Barges Ships Ferries Average ton-miles per gallon of fuel 186 146 19 Average fuel costs per ton-mile $0.016 $0.021 $2.37 CO2 emissions per ton mile (kilograms) 0.05 0.08 0.53 Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

Table 19. Comparison of fuel use and costs per ton-mile for rail and trucking, 2007–2010, 2011$

Railroad Trucks Average miles per gallon 0.1 4.5 Average ton-miles per gallon of fuel 280 48 Average fuel costs per mile $21 $0.61 Average fuel costs per ton-mile $0.03 $0.27 Sources: U.S. DOT, BTS; U.S. DOE, EIA; AKRR; company proprietary information; author calculations.

Table 20. Comparison of fuel use and costs per passenger-mile for rail, ferry and air, 2007–2010, 2011$

Railroad Ferry Air* Average miles per gallon 0.2 0.06 -- Average passengers-miles per gallon of fuel 120 12 50 Average fuel costs per mile $15 $45 -- Average fuel costs per passengers-mile $0.02 $3.83 $0.09 *U.S. average for comparison purposes only. Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Army Corps of Engineers, Waterborne Statistics; AMHS; AKRR; IFA; Ingram, 2008; company proprietary information; author calculations.

Figure 14. Comparison of fuel use and costs per ton-mile for Alaska transportation, 2007–2010, 2011$ Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

*U.S. average for comparison purposes only. Figure 15. Comparison of fuel use and costs per passenger-mile for rail and air, 2007–2010, 2011$ Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Army Corps of Engineers, Waterborne Statistics; AMHS; AKRR; IFA; Ingram, 2008; company proprietary information; author calculations.

Change in Transportation Costs and Impact on Alaska Industries This section examines the economic impact of changes in oil prices and refined petroleum fuel prices on the Alaska economy, industries, and households, and the resulting effect of those changes on transportation services. We initiate the discussion with a description of the Alaska transportation sectors analyzed—air, water, truck, and rail—in employment, payroll, and the number of firms in 2008 (Table 21) and compare these characteristics and changes with data from 2010 (Table 22 and Table 23).

Rail transportation does not appear in these employment numbers because Alaska Railroad employees are state employees; those numbers are part of the state government sector.

Table 21. Alaska transportation sector employment, payroll, and firms by transportation mode, 2008

Payroll ($1,000) Total NAICS Transportation mode Employees1 1st quarter Annual firms code 48---- Transportation and warehousing 19,956 268,504 1,181,652 1,133 481 Air transportation 6,858 89,340 392,103 221 48111 Scheduled 5,895 78,191 329,799 111 481111 Scheduled passenger 4,522 58,235 251,374 79 Scheduled freight air 481112 1,373 19,956 78,425 32 transportation 48121 Nonscheduled 963 11,149 62,304 110 Nonscheduled chartered 481211 628 5,975 37,806 93 passenger 481212 Nonscheduled chartered freight c D D 6 481219 Other nonscheduled 137 2,120 8,410 11 483 Water transportation 1,009 12,422 55,312 82 48311 Deep sea, coastal, and great lakes 987 11,982 53,492 58 483111 Deep sea freight transportation e D D 3 483112 Deep sea passenger a D D 1 483113 Coastal and great lakes freight f 7,461 33,888 49 483114 Coastal and great lakes passenger 4 S 165 5 48321 Inland water b 440 1,820 24 483211 Inland water freight a D D 9 483212 Inland water passenger b 424 1,210 15 484 Truck transportation 2,969 36,958 165,170 229 1Paid employees for pay period including March 12 (number) D: withheld to protect privacy (for quarterly/annual) a: 0–19; b: 20–99; c: 100–249; e: 250–499; f: 500–999; g: 1000–2499; h: 2500–4999 Source: U.S. Census, http://censtats.census.gov/cgi-bin/cbpnaic/cbpdetl.pl

Table 22. Alaska transportation sector employment, payroll, and firms by transportation mode, 2010

Payroll ($1,000) Total NAICS Transportation mode Employees1 1st quarter Annual firms code 48---- Transportation and warehousing 17,974 256,733 1,118,746 1,113 481 Air transportation 5,845 75,818 320,111 207 48111 Scheduled 4,901 64,742 257,983 102 481111 Scheduled passenger 4,106 49,940 201,292 71 Scheduled freight air 481112 795 14,802 56,691 31 transportation 48121 Nonscheduled 944 11,076 62,128 105 Nonscheduled chartered 481211 658 6,487 37,586 87 passenger 481212 Nonscheduled chartered freight c D D 6 481219 Other nonscheduled 95 1,429 8,151 12 483 Water transportation 1,312 14,754 70,749 93 48311 Deep sea, coastal, and great lakes 1,289 14,594 67,730 64 483111 Deep sea freight transportation c D D 3 483113 Coastal and great lakes freight g D D 55 483114 Coastal and great lakes passenger a 17 169 6 48321 Inland water 23 S 3,019 29 483211 Inland water freight a D D 12 483212 Inland water passenger a 71 1,055 17 484 Truck transportation 2,530 28,590 136,415 225 1Paid employees for pay period including March 12 (number) D: withheld to protect privacy (for quarterly/annual) a: 0–19; b: 20–99; c: 100–249; e: 250–499; f: 500–999; g: 1000–2499; h: 2500–4999 Source: U.S. Census, http://censtats.census.gov/cgi-bin/cbpnaic/cbpdetl.pl

Table 23. Change in Alaska transport sector employment, payroll, and firms, 2008–2010

Payroll ($1,000) Total NA ICS Transportation mode Employees1 1st quarter Annual firms code 48---- Transportation and warehousing -9.9% -4.4% -5.3% -1.8% 481 Air transportation -14.8% -15.1% -18.4% -6.3% 48111 Scheduled -16.9% -17.2% -21.8% -8.1% 481111 Scheduled passenger -9.2% -14.2% -19.9% -10.1% Scheduled freight air 481112 -42.1% -25.8% -27.7% -3.1% transportation 48121 Nonscheduled -2.0% -0.7% -0.3% -4.6% 481211 Nonscheduled chartered passenger 4.8% 8.6% -0.6% -6.5% 481212 Nonscheduled chartered freight NA NA NA 0.0% 481219 Other nonscheduled -30.7% -32.6% -3.1% 9.1% 483 Water transportation 30.0% 18.8% 27.9% 13.4% 48311 Deep sea, coastal, and great lakes 30.6% 21.8% 26.6% 10.3% 483111 Deep sea freight transportation NA NA NA -100.0% 483112 Deep sea passenger NA NA NA 200.0% 483113 Coastal and great lakes freight NA NA NA 12.2% 483114 Coastal and great lakes passenger NA NA 2.4% 20.0% 48321 Inland water 65.9% 20.8%

483211 Inland water freight NA NA NA 33.3% 483212 Inland water passenger NA -83.3% -12.8% 13.3% 484 Truck transportation -14.8% -22.6% -17.4% -1.8% Source: U.S. Census, http://censtats.census.gov/cgi-bin/cbpnaic/cbpdetl.pl

Table 23 shows that between 2008 and 2010, employment in the transportation sector decreased by approximately 10%, with air transportation leading the decline at 15%. Most of the air decline was concentrated in scheduled freight transportation. The annual payroll decline in air transportation was 18%, again led by declines in scheduled freight transportation.

In contrast, employment in water transportation increased 30%, and trucking employment declined approximately 15%. These numbers reflect that the most energy-intensive transportation sectors experienced the largest decrease in employment. However, transportation services were also affected by the recession in addition to changes in crude oil prices.

Besides declines in employment and payroll, the number of firms decreased by 20, mostly in air transportation (Table 24 and Table 25). However, the number of water transportation firms increased, primarily in the smallest firm-size category. The increase in water transportation firms or businesses is likely the source of the employment increase shown in Table 27. Table 24 and Table 25 show the total number of firms, but not all firm-size classifications.

Table 24. Number of transportation industry firms by size, 2008*

# of firms by employment-size class

Total Industry Industry code description # '1-4' '5-9' '10-19' code 48---- Transportation and warehousing 1,133 646 175 132 481 Air transportation 221 96 34 35 48111 Scheduled 111 23 16 25 481111 Scheduled passenger 79 16 7 19 481112 Scheduled freight air transportation 32 7 9 6 48121 Nonscheduled 110 73 18 10 481211 Nonscheduled chartered passenger 93 62 16 8 481212 Nonscheduled chartered freight 6 3 1 1 481219 Other nonscheduled 11 8 1 1 483 Water transportation 82 55 11 8 48311 Deep sea, coastal, and great lakes 58 32 11 7 483111 Deep sea freight transportation 3 1 0 0 483112 Deep sea passenger 1 1 0 0 483113 Coastal and great lakes freight 49 25 11 7 483114 Coastal and great lakes passenger 5 5 0 0 48321 Inland water 24 23 0 1 483211 Inland water freight 9 9 0 0 483212 Inland water passenger 15 14 0 1 484 Truck transportation 229 130 33 28 *Not a complete list of all size firms, but the total includes all firms of all sizes. Source: U.S. Census, http://censtats.census.gov/cgi-bin/cbpnaic/cbpdetl.pl

Table 25. Number of transportation industry firms by size, 2010*

# of firms by employment-size class Total Industry Industry code description # '1-4' '5-9' '10-19' code 48---- Transportation and warehousing 1,113 655 160 121 481 Air transportation 207 79 44 33 48111 Scheduled 102 20 23 17 481111 Scheduled passenger 71 10 10 15 481112 Scheduled freight air transportation 31 10 13 2 48121 Nonscheduled 105 59 21 16 481211 Nonscheduled chartered passenger 87 50 17 14 481212 Nonscheduled chartered freight 6 2 2 1 481219 Other nonscheduled 12 7 2 1 483 Water transportation 93 67 9 6 48311 Deep sea, coastal, and great lakes 64 39 8 6 483111 Deep sea freight transportation 3 1 1 0 483113 Coastal and great lakes freight 55 32 7 6 483114 Coastal and great lakes passenger 6 6 0 0 48321 Inland water 29 28 1 0 483211 Inland water freight 12 12 0 0 483212 Inland water passenger 17 16 1 0 484 Truck transportation 225 133 28 25 *Not a complete list of all size firms, but the total includes all firms of all sizes. Source: U.S. Census, http://censtats.census.gov/cgi-bin/cbpnaic/cbpdetl.pl

Using IMPLAN, we examine the sensitivity of the Alaska economy to the increase in transportation costs and fuel prices. We do this by passing cost increases to transportation consumers, which are households and industries that purchase transportation services. The input-output modeling methodology described earlier informs us regarding the transportation intensity of each sector. In input-output terminology, the model gives us the gross absorption of each of the transportation modes, which is simply the cents spent on that activity per dollar of output. Given that we also have the total output produced by each sector, we know the dollar amount that is spent by any given industry on transportation services. In the very short run, all else held constant, we simply analyze how the increase in transportation service prices would influence the sector’s demand for inputs, which in this setting is equivalent to a decrease in value added, given that output is kept constant (see Appendix D for a glossary of economic impact terms).

Price increases are applied in the following way: We multiply the transportation sector’s refined petroleum fuel-price increase times the share that refined petroleum fuel purchases represent per dollar of output for that transportation sector (Table 26). For example, if refined petroleum fuel products are 0.312 of each unit of output of air transportation, then a 29.16% increase in the price of refined petroleum fuel products results in a 9.1% increase in the cost of a unit of air transportation or:

0.312 * 0.2916 = 0.091

Thus, in this example of air transport, a 29.16% increase in jet fuel prices translates to a 0.091 increase in input costs for each unit of air transportation services output, or essentially an approximately 9% increase in costs. Table 26 shows these fuel price increases for each transportation mode based on U.S. EIA reported price increases between 2008 and 2010. We use these price increases in our impact simulation. The actual increase for each transportation subsector depends on the amount of fuel purchased. Table 26 shows these cost increases for all four subsectors.

Table 26. Alaska transportation services estimated price increase, 2008–2010 Fuel price increase Estimated price Gross absorption coefficient (dollars

percentage (EIA) increase spent on petroleum per dollar of output)

Air 29.16% 9.1% 0.312

Rail 29.17% 3.4% 0.131

Water 27.74% 7.4% 0.266

Truck 26.17% 3.9% 0.149

Sources: IMPLAN Input-Output Model; U.S. DOE, EIA; author calculations.

Before presenting the results, we show a list of the most transport-intensive industries (Table 27). In other words, these industries spend the largest portions per dollar of output on transportation services expenditures. The columns represent the cents per dollar of output being spent on these transport categories. We also show the transport share of the budget, which is simply the cents that are spent on transport as defined by the four sectors. Given that the Alaska economy had 257 sectors in 2008, we only list those that are the heaviest users in relation to their operation size. This list clearly does not show which industries are the biggest consumers of transportation services in absolute terms.

Table 27. Transportation usage intensity, 2008

Transport Industry name Air Rail Water Truck share Ready-mix concrete manufacturing 0.0029 0.0164 0.0032 0.1001 0.1227 Cut stone and stone product manufacturing 0.0051 0.0160 0.0027 0.0526 0.0764 Fertilizer manufacturing 0.0003 0.0111 0.0018 0.0607 0.0740 Office supplies (except paper) manufacturing 0.0019 0.0098 0.0009 0.0415 0.0541 Coffee and tea manufacturing 0.0033 0.0023 0.0035 0.0444 0.0536 Other animal food manufacturing 0.0011 0.0232 0.0039 0.0241 0.0523 Other pressed and blown glass and glassware manuf. 0.0020 0.0264 0.0018 0.0221 0.0523 Transport by truck 0.0025 0.0098 0.0008 0.0388 0.0518 Other aircraft parts and auxiliary equipment manuf. 0.0030 0.0056 0.0009 0.0399 0.0494 Scientific research and development services 0.0026 0.0010 0.0415 0.0032 0.0483 Mining coal 0.0002 0.0303 0.0021 0.0149 0.0475 Breweries 0.0011 0.0163 0.0007 0.0291 0.0472 Dog and cat food manufacturing 0.0007 0.0195 0.0022 0.0218 0.0442 Wood windows and doors and millwork manuf. 0.0027 0.0123 0.0001 0.0280 0.0430 Engineered wood member and truss manuf. 0.0034 0.0140 - 0.0254 0.0428 Mining and quarrying other nonmetallic minerals 0.0004 0.0096 0.0004 0.0302 0.0405 Jewelry and silverware manufacturing 0.0046 0.0016 0.0014 0.0328 0.0404 Snack food manufacturing 0.0016 0.0080 0.0022 0.0284 0.0401 Animal slaughtering and processing 0.0034 0.0010 0.0001 0.0355 0.0400 All other food manufacturing 0.0028 0.0069 0.0014 0.0274 0.0386 Mining and quarrying stone 0.0005 0.0040 0.0017 0.0323 0.0386 Fluid milk and butter manufacturing 0.0020 0.0016 0.0001 0.0343 0.0380 Concrete pipe, brick, and block manufacturing 0.0029 0.0096 0.0004 0.0240 0.0369 Sawmills and wood preservation 0.0010 0.0090 0.0000 0.0259 0.0360 US Postal Service 0.0130 0.0084 0.0059 0.0084 0.0357 Wineries 0.0020 0.0023 0.0016 0.0291 0.0351 Prefabricated wood building manuf. 0.0031 0.0082 - 0.0202 0.0314 Wood container and pallet manuf. 0.0028 0.0084 0.0002 0.0181 0.0295 Source: IMPLAN Input-Output Model, author calculations.

To apply the price shock, we simply increase each industry’s expenditures on the respective transportation services by the percentages indicated in Table 26. Given that the various transportation services face different price increases, the cost increase faced by any given industry will depend upon the specific composition of the transportation services it consumes. We list the most affected industries in Table 28. We list these industries sorted according to the ones incurring the largest absolute increase in terms of their intermediate inputs expenditures on transportation services. We present the information in Table 28 in terms of value added lost as a result of the increase in the cost of transportation services that the industry purchases. The two presentations are equivalent given that:

Intermediate Inputs + Value Added = Total Output

Table 28. Value-added losses (most affected industries in absolute terms), millions$$

Industry Value added Refined petroleum Value as share of Code Name added New Decrease inputs 332 Transport by air 670.0 517.9 -152.1 31% 115 Petroleum refineries 120.0 57.5 -62.5 8% 17 Commercial fishing 110.0 70.6 -39.4 32% 334 Transport by water 160.0 121.2 -38.8 27% 337 Transport by pipeline 240.0 208.9 -31.1 21% 335 Transport by truck 300.0 277.7 -22.3 15% 339 Couriers and messengers 470.0 448.6 -21.4 12% 34 Construction of new commercial/health care 710.0 695.1 -14.9 4% 36 Construction of other new nonresidential structures 410.0 395.8 -14.2 6% 388 Services to buildings and dwellings 100.0 88.1 -11.9 22% 37 Construction of new residential housing 400.0 393.6 -6.4 3% 20 Extraction of oil and natural gas 2,000.0 1,994.0 -6.0 1% 319 Wholesale trade businesses 750.0 745.5 -4.5 1% 29 Support activities for oil and gas operations 1,500.0 1,495.8 -4.2 1% 39 Maintenance and repair of nonresidential structures 250.0 245.9 -4.1 4% 351 Telecommunications 1,000.0 996.0 -4.0 1% 333 Transport by rail 42.9 39.0 -3.8 14% 31 Electric power generation, transmission, distribution 640.0 636.7 -3.3 1% 369 Architectural, engineering, and related services 680.0 676.9 -3.1 1% 336 Transit and ground passenger transportation 65.0 62.0 -3.0 11% 61 Seafood product preparation and packaging 450.0 447.1 -2.9 0% 413 Food services and drinking places 790.0 787.2 -2.8 1% Subtotals and share of total 11,857.9 11,401.1 -456.8 92% Source: IMPLAN Input-Output Model, author calculations.

The total of value added lost because of the transportation cost increase is $456.8 million. This loss in value added is equivalent to the additional cost of commodity inputs (in this case, higher transportation services). This presentation gives us a sense of the value-added losses that would be incurred by the industries that are the largest consumers of transportation services in absolute terms. This picture, however, does not provide us with information regarding the transportation intensity of these sectors.

An alternative way of examining the shock is to investigate the share of the industry’s value added that would be lost as a result of the transportation cost increase. Table 29 lists the most sensitive industries; it shows the relative decrease in the sector’s value added due to the transportation cost increases.

Note that the extent to which Alaska commodities, especially raw materials, are extracted and exported by foreign companies using foreign shipping fleets affects the calculation. These transportation service values are not included in this exercise because such transactions are not part of the Alaska economy. For example, if a Canadian mining company extracts ore and ships it to Canada or Asia for smelting, the

payment for the ore, including the cost of shipping, does not enter the Alaska economy. Instead, the payment is received in Canada and reflected in Canadian regional economic accounting. If the payment for shipping is a separate transaction but paid to a foreign shipper, this transaction similarly would not be reflected in the Alaska regional economy. To the extent that shippers purchase fuel in Alaska, these fuel purchase transactions would be part of the Alaska regional economy as well as the emissions resulting from the fuel purchased in Alaska.

Table 29. Value-added losses (most affected industries relative to own value added)

Industry Value Industry name code added

66 Coffee and tea manufacturing -3.40% 130 Fertilizer manufacturing -2.49% 42 Other animal food manufacturing -2.26% 91 Apparel accessories and other apparel manufacturing -2.06% 347 Sound recording industries -2.01% 59 Animal (except poultry) slaughtering and processing -1.97% 72 Wineries -1.92% 310 Jewelry and silverware manufacturing -1.76% 161 Ready-mix concrete manufacturing -1.75% 70 Soft drink and ice manufacturing -1.62% 138 Soap and cleaning compound manufacturing -1.53% 65 Snack food manufacturing -1.52% 71 Breweries -1.33% Source: IMPLAN Input-Output Model, author calculations.

This exercise examines the very short-run implications of a price increase in transportation services if the consuming industries were to continue using the same amounts of commodity input and no other behavioral responses were to occur. It is clear the sensitivity varies greatly from one sector to the next and that different metrics yield different rankings.

To assess the impact of these increases on actual usage, we look at what happened to transportation usage by mode between 2008 and 2010. Before looking at usage of the different industries, we examine an aggregate picture (Table 30).

Table 30. Change in institution and industry commodity demand for transportation services between 2008 and 2010

Commodity 2008 2010 2008$ % Change Demand Rail

Institution $141,582,592 $144,231,376 $136,582,742 -4% Industry $76,224,520 $61,473,528 $58,213,568 -24% All $217,807,112 $205,704,904 $194,796,311 -11%

Air

Institution $1,855,906,304 $1,521,707,008 $1,481,701,079 -20% Industry $92,350,312 $94,027,944 $91,555,934 -1% All $1,948,256,616 $1,615,734,952 $1,573,257,013 -19%

Water

Institution $523,841,536 $510,785,344 $458,103,448 -13% Industry $25,166,378 $6,796,376 $6,095,404 -76% All $549,007,914 $517,581,720 $464,198,852 -15%

Truck

Institution $441,049,024 $356,745,728 $348,725,052 -21% Industry $317,363,488 $267,229,648 $261,221,552 -18% All $758,412,512 $623,975,376 $609,946,604 -20% Institution demand, or final demand as it is sometimes called, is demand for goods and services for final use. Final use means that the goods or service will be consumed and not incorporated into another product. In contrast, industry commodity demand is the intermediate use of a good or service to produce another good or service. An example would be the tourism industry purchasing air transportation services as part of a tour package. Commodity demand is the amount of purchases of transportation services by institutions (final purchases) and industry (intermediate purchases as inputs to production). Source: IMPLAN Input-Output Model for Alaska (2008–2010) and author calculations.

The aggregate picture presented in Table 30 makes clear that the demand for transportation decreased along with economic conditions, including the major recession. However, the magnitude of the decrease is not consistent across the four sectors. Also, institutional demand decreased for transport by air, while demand by industries was flat; on the other hand, we see that demand by industries for transportation by water has taken a sharp decrease. However, industry commodity demand for water transportation is only 5% of total water transportation demand.

Next, we examine how transportation-related expenditures have changed for the different sectors.

Using IMPLAN data for 2008 and 2010, we look at how expenditures on different transportation modes shifted for a given industry by following these steps:  First, we generate transportation dollars for years 2008 and 2010, which are comprised of the expenditures made by the different sectors on air, water, rail, and truck transportation services.

 Second, we generate shares that these different transportation sectors comprise from the overall transport dollars.  Third, we compare how the allocation of transportation dollars shifted between 2008 and 2010.

By comparing the shares across years, we get an idea, for example, if air transportation represents a bigger or smaller share of the overall transportation dollars in the Alaska economy.

Before showing the changes, we list the industries with the highest absolute transportation expenditures and the share those expenditures represent of their total output (Table 31).

Table 31. Alaska industries with largest transportation expenditure inputs, 2008, million$$

Industry Code Name Output Transport Share 61 Seafood product preparation and packaging $3.3 $67.3 0.0204 29 Support activities for oil and gas operations $2.8 $46.0 0.0164 335 Transport by truck $570.0 $29.5 0.0518 28 Drilling oil and gas wells $920.0 $18.2 0.0198 Construction of new nonresidential commercial and 34 $1,500.0 $17.4 0.0116 health care structures 37 Construction of new residential single-/multi-family $870.0 $16.1 0.0185 31 Electric power generation, transmission, distribution $840.0 $14.8 0.0176 24 Mining gold, silver, and other metal ore $1,200.0 $13.2 0.0110 413 Food services and drinking places $1,500.0 $12.1 0.0081 432 Other state and local government enterprises $1,100.0 $11.9 0.0108 36 Construction of other new nonresidential structures $860.0 $11.0 0.0128 351 Telecommunications $2,300.0 $10.2 0.0044 71 Breweries $220.0 $10.2 0.0464 161 Ready-mix concrete manufacturing $81.0 $9,.9 0.1227 332 Transport by air $1,800.0 $8,.5 0.0047 115 Petroleum refineries $3,100.0 $8.2 0.0026 Source: IMPLAN Input-Output Model, author calculations.

Table 32 shows the mode shifts between 2008 and 2010 for sectors with the highest transport expenditures. The changes are calculated using the following formula:

Change Air = (ShareAir2010 –ShareAir2008)/ (ShareAir2008)

Table 32. Mode shifts for Alaska industries with largest transportation expenditure inputs following imposed fuel price increases

Change Industry Name Code Air Rail Water Truck Seafood product preparation and packaging 61 0.036 -0.017 None -0.109 Support activities for oil and gas operations 29 0.012 -0.041 -0.748 -0.337 Transport by truck 335 0.090 0.033 -0.729 -0.145 Construction of other new nonresidential structures 36 0.108 0.051 -0.724 1.129 Other state and local government enterprises 432 0.391 0.319 -0.656 0.003 Food services and drinking places 413 0.041 -0.014 -0.74 -0.081 Petroleum refineries 115 0.995 0.885 -0.506 0.531 Electric power generation, transmission, distribution 31 -0.207 -0.249 -0.804 -0.355 Mining gold, silver, and other metal ore 24 -0.143 -0.191 -0.788 -0.240 Maintenance/repair/construction nonresidential 39 0.401 0.328 -0.652 0.870 structures Construction of other new residential structures 38 0.046 -0.008 -0.74 1.136 Monetary/depository credit intermediation activities 354 0.945 0.818 -0.523 1.532 Offices of physicians, dentists, other health practitioners 394 0.260 0.193 -0.691 0.293 Ready-mix concrete manufacturing 161 0.105 0.047 -0.726 -0.277 Telecommunications 351 -0.214 -0.254 -0.807 -0.357 Construction of new nonresidential commercial and 34 -0.043 -0.093 -0.764 -0.618 health care structures A complete listing of all the industries mode shifts is available upon request. Source: IMPLAN Input-Output Model, author calculations.

In looking at this information, it is important to note that the demand for any given commodity (for example, water transportation) is comprised of industrial and institutional consumption. Institutional consumption (or final consumption as opposed to industry consumption for the production of a final product) includes nine different household income groups, government, and domestic and foreign exports.

Within demand for water transportation, the pattern was not consistent. Industrial demand declined (-76%). The change in institutional demand was more complicated, with the demand for exports to foreign countries increasing considerably and the demand for exports to the U.S. markets declining similarly. Further analysis shows that industrial demand for water transportation decreased significantly. Table 33 shows a breakdown of transport demand between industry and institutions; it also shows the demand for water transportation in both years (2008 and 2010) and how the subcomponents of institutional demand changed. We are uncertain of the drivers of these demand shifts, but note that most smelting of ores extracted from Alaska occurs in export locations, and mineral prices continued to increase during this period. In addition, domestic oil shipments from Valdez to refineries in the Lower 48 decreased by four million short tons between 2008 and 2010, reflecting the continued reduction in North Slope oil production (U.S. Army Corps of Engineers, 2012a, b).

Table 33. Institution and industry demand for water transportation, 2008 and 2010 (2008$)

2008 2010 % change Institution Demand $523.8 $458.1 -13% Foreign Exports $191.1 $335.1 75% Domestic Exports $275.8 $92.3 -67% Industry Demand $25.2 $6.1 -76% Total Demand $547.9 $464.1 -15% Source: IMPLAN Input-Output Model, author calculations.

Impact of Change in Transportation Costs on Alaska Households We apply the same transport price increases to households as we do to industry—by multiplying the mode's price increase times the share that refined petroleum purchases represent per dollar of output (see Table 26 for details). Nine household income categories are in our model, and they all consume transportation services, which now cost more as a result of the price shock imposed; this is the same price increase faced by industries, as discussed earlier. The changes for households reflect direct and indirect consumer purchases of transportation services. For example, an apple purchased at a grocery store includes imbedded transportation services that include shipping from its origin in the orchard to the grocery store shelves. The scenario presented in Table 34 assumes that these household groups will continue consuming the same amounts of transportation services at their higher price to infer how much “income loss” would be incurred after the price increases. Transportation dollars are once again comprised of dollars spent on air, rail, truck, and water transportation services (Table 34).

Table 34. Increases in the cost of transportation services to households by income groups after an increase in the price of refined petroleum products, millions$$

Direct household expenditures on transportation services All households with Air Rail Water Truck incomes of: Before After Before After Before After Before After less than $10,000 $6.7 $7.3 $0.1 $0.1 $0.1 $0.1 $3.8 $3.9 $10,000-$14,999 $3.5 $3.9 $0.3 $0.3 $0.2 $0.2 $3.0 $3.1 $15,000-$24,999 $8.9 $9.7 $1.3 $1.3 $0.7 $0.7 $10.7 $11.1 $25,000-$34,999 $10.7 $11.7 $1.4 $1.4 $0.7 $0.8 $9.8 $10.2 $35,000-$49,999 $23.0 $25.0 $1.2 $1.3 $1.3 $1.4 $19.6 $20.3 $50,000-$74,999 $36.0 $39.3 $3.7 $3.9 $4.1 $4.4 $28.8 $29.9 $75,000-$99,999 $27.9 $30.5 $2.3 $2.4 $2.5 $2.6 $27.1 $28.2 $100,000-$149,999 $38.4 $41.9 $2.8 $2.9 $3.0 $3.2 $27.9 $29.0 $150,000+ $47.1 $51.4 $3.5 $3.6 $3.5 $3.7 $38.5 $40.0 Total $202.3 $220.8 $16.7 $17.2 $16.0 $17.2 $169.1 $175.7 Increase cost to households $18.4 $0.6 $1.2 $6.6 Source: IMPLAN Input-Output Model, author calculations.

The total effect on households would be equivalent to a loss of income equaling $26.8 million (Table 35). This “loss” is simply the first-order decrease in income available for household purchases as a result of

the increase in prices of transportation services. The table shows that among households at almost every income level, the increase in spending for transportation services after the imposed price increase is approximately the same percentage (6% to 7%).

Table 35. Increases in the cost of air, rail, water, and truck transportation services to households by income groups after an increase in the price of refined petroleum products, millions$$

All households Direct household expenditures on transportation services with incomes of: Before After Difference Share of change less than $10,000 $10.7 $11.5 $0.8 3% $10,000-$14,999 $7.0 $7.4 $0.5 2% $15,000-$24,999 $21.6 $22.9 $1.3 5% $25,000-$34,999 $22.6 $24.1 $1.5 5% $35,000-$49,999 $45.0 $48.0 $3.0 11% $50,000-$74,999 $72.6 $77.5 $4.8 18% $75,000-$99,999 $59.8 $63.7 $3.9 14% $100,000-$149,999 $72.1 $77.0 $4.9 18% $150,000+ $92.6 $98.7 $6.2 23% Total $404.1 $430.8 $26.8 Source: IMPLAN Input-Output Model, author calculations.

The increases in transportation service costs or equivalent “household income losses” in Table 34 depict the assumption that all households continue to consume the same level of transportation services even after the price increases. Table 35 shows two types of information: (1) how much more those services would cost based on the transportation-services consumption patterns of specific household income groups, and (2) the share of increase in cost of transportation services that each household income group would absorb of the total cost increases. Table 34 shows how these increased costs would be spread across the air, rail, water, and trucking sectors. Again, these figures are the total amount each household income group would pay for transportation services by mode following increases in the cost of these services and assuming no change in rate of consumption. The cost increases result from an increase in the price of refined petroleum products.

In contrast, Table 36 shows how the allocation of transportation dollars would shift for different household income groups if we remove the assumption that they will continue to consume transportation services at the same rate as before the price increase. This exercise simply examines how the shares of each sector change because of price increases imposed for transportation services; it reflects actual cost changes between 2008 and 2010. The changes in Table 36 also include the reduction in consumer purchasing that resulted from the recession.

Table 36. Shift in household purchases of air, rail, water, and truck transportation services due to increases in the cost of transportation services

All households with incomes of: Air Rail Water Truck less than $10,000 14.8% -35.3% -43.3% -23.4% $10,000-$14,999 20.9% -31.9% -40.3% -19.3% $15,000-$24,999 27.9% -44.6% -51.4% -14.6% $25,000-$34,999 23.0% -30.7% -39.2% -17.9% $35,000-$49,999 20.6% -32.1% -40.4% -19.5% $50,000-$74,999 22.4% -31.1% -39.5% -18.4% $75,000-$99,999 23.2% -30.6% -39.1% -17.8% $100,000-$149,999 20.0% -32.4% -40.7% -20.0% $150,000+ 21.0% -31.8% -40.2% -19.2% Source: IMPLAN Input-Output Model, author calculations.

In Table 37 we show how much more households, by income groups, would pay directly for refined petroleum products (primarily diesel and gasoline) due to fuel price increases if we assume that they continue to consume at the same levels before and after the price increases. These dollar amounts are the increases in household expenditures for refined petroleum products purchased by the household directly, such as gasoline for personal vehicles, rather than transportation services. Utilities are not included, so the calculation does not encompass things such as natural gas for space heating or electricity, or diesel to produce electricity. Under this scenario related to direct household income group purchases of refined petroleum products, the increase in costs to households, or the equivalent loss of income, amounts to approximately $124 million.

Table 37. Increases in household expenditures due to an increase in the cost of refined fuel prices, millions$$

Direct household expenditures on refined petroleum products Share of: Household All households Expenditures Statewide budget with incomes of: Before After Difference change Before After less than $10,000 $10.4 $13.3 $2.9 2% 2.2% 2.8% $10,000-$14,999 $8.9 $11.4 $2.5 2% 1.9% 2.4% $15,000-$24,999 $28.9 $36.8 $7.9 6% 2.5% 2.4% $25,000-$34,999 $31.4 $40.1 $8.7 7% 2.0% 2.6% $35,000-$49,999 $52.4 $66.9 $14.5 12% 1.9% 2.4% $50,000-$74,999 $99.5 $127.1 $27.6 22% 1.9% 2.4% $75,000-$99,999 $75.3 $96.1 $20.8 17% 1.8% 2.3% $100,000-$149,999 $81.6 $104.3 $22.7 18% 1.8% 2.2% $150,000+ $59.5 $76.0 $16.5 13% 1.4% 1.7% Total $447.9 $572.0 $124.1 Source: IMPLAN Input-Output Model, author calculations.

Carbon Emissions Analysis Because a carbon emissions tax has a similar effect as a price increase, we also assessed the potential effects of such a tax. To explore the potential impact of an emissions tax on transportation sector output and employment, we first examine the direct emissions from fuel use for the various transportation modes and then analyze the potential impact of an emissions tax.

Direct Emissions from Fuel Use Table 38 and Figure 16 provide an estimate of CO₂ emissions by mode for the years 2007 to 2010. These emissions estimates are solely for the direct fuel used by the various modes and do not include fuel and emissions in ports, rail yards, or airports. The emissions estimates also do not include “imbedded” fuel use and emissions by the industry sector as a whole, such as heating and lighting of buildings and other energy uses. As a result, these emissions estimates are a subset of our industry analysis of emissions and the potential effects of an emissions tax.

Millions Millions tons CO2 6

- 2007 2008 2009 2010 Air Water Trucks Rail

Figure 16. Comparison of annual emissions by transportation sector, intrastate and exiting only, 2007–2010 Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

Table 38. Estimated aviation, shipping, trucking, and rail emissions, Alaska 2007–2010

Aviation Scheduled intra-state 2007 2008 2009 2010 gallons 30,684,229 31,457,698 27,330,122 27,846,200

CO2 emissions [mt] 293,587 300,987 261,495 266,432 Non-scheduled intra-state gallons 5,963,125 9,301,368 6,732,250 7,859,375

CO2 emissions [mt] 57,055 88,995 64,414 75,199 Exiting (scheduled & non-scheduled) gallons 1,193,994,646 967,954,934 762,383,628 925,632,425

CO2 emissions [mt] 11,411,037 9,250,826 7,283,021 8,847,839 Entering (scheduled & non-scheduled) gallons 1,207,421,777 982,127,718 772,386,783 934,194,928

CO2 emissions [mt] 11,539,361 9,386,276 7,378,581 8,929,685 Aviation totalsgallons 2,438,063,777 1,990,841,718 1,568,832,783 1,895,532,928

CO2 emissions [mt] 23,301,040 19,027,084 14,987,511 18,119,155

Alaska CO2 emissions [mt] 11,761,679 9,640,808 7,608,930 9,189,470 Water Barges/Ships gallons 95,383,724 94,056,619 95,190,267 91,307,812 gal. (intra-state & exiting) 78,243,753 77,260,621 78,012,015 73,839,840

CO2 emissions [mt] 825,389 814,822 823,256 781,901 Ferries gal. (intra-state & exiting) 11,946,490 10,455,345 10,150,310 10,450,611

CO2 emissions [mt] 109,558 97,318 93,304 96,065 Total Water gallons 107,330,214 104,511,963 105,340,577 101,758,423

CO2 emissions [mt] 934,947 912,140 916,560 877,966 Trucks (only one year of data) gallons 9,800,000 9,800,000 9,800,000 9,786,844 gallons (intra-state & exiting) 8,900,000 8,900,000 8,900,000 8,805,455

CO2 emissions [mt] 90,335 90,335 90,335 89,375 Railroad gallons (freight) 4,342,932 3,241,907 3,618,137 3,993,404

CO2 emissions [mt] 44,081 32,905 36,724 40,533 gallons (passengers) 1,220,758 1,228,142 1,195,411 1,149,241

CO2 emissions [mt] 12,391 12,466 12,133 11,665 Total rail gallons 5,563,690 4,470,049 4,813,548 5,142,645

CO2 emissions [mt] 56,471 45,371 48,858 52,198 Total Alaska CO2 emissions [mt] 12,843,433 10,688,654 8,664,682 10,209,009 Notes: Numbers in black are directly from the BTS and EIA data. Blue numbers were calculated from BTS, EIA data and other proprietary data sources. Total weight includes freight, mail, and passengers. Aircraft type numbers are T100 data codes. Fuel prices in blue are estimates based on OPIS price data and limited proprietary data provided. Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; AKRR; IFA; company proprietary information; author calculations.

Despite this lack of precision, we can see that CO₂ emissions from barges and ships range from about 10% to 15% per ton-mile of freight carried, as compared with ferries (Table 39). The Alaska Railroad emissions are about 20% of that from trucks, per ton-mile transported (Table 40). The railroad moves passengers at approximately half the per passenger-mile emissions of the national estimate for aviation and about 10% of the estimate for ferries (Table 41). The per passenger-mile carbon emissions of ferry transportation is, on average, about five times greater than the national estimate for air passenger

transportation (without an aviation emissions multiplier), likely as a result of the Alaska Marine Highway System’s low capacity factor and passenger-miles per gallon of fuel. Emission multipliers are used for aviation because the impacts of emissions are greater at altitude (Jardine, 2009). These emissions multipliers range in value from 1.9 to 4, depending on the emissions calculator and studies (Jardine, 2009). Use of an emissions multiplier reduces this difference, but air transportation emissions per passenger-mile still remain below the transportation emissions of ferries (Table 41). Estimates of aviation emissions with and without an emissions multiplier are shown in Table 41.

Table 39. Estimated water transportation ton-miles per gallon and fuel costs and emissions per ton-mile transported, Alaska 2007–2010, 2011$

Water Barges Ships Ferries Average ton-miles per gallon of fuel 186 146 19 Average fuel costs per ton-mile $0.016 $0.021 $2.37

CO2 emissions per ton-mile (kilograms) 0.05 0.08 0.53 Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Waterborne Statistics; AMHS; IIFA; company proprietary information; author calculations.

Table 40. Comparison of estimated land transportation modes ton-miles per gallon and fuel costs and emissions per ton-mile transported, Alaska 2007–2010, 2011$

Railroad Trucks Average ton-miles per gallon of fuel 279 48 Average fuel costs per ton-mile $0.01 $0.06

CO2 emissions per ton-mile (kilograms) 0.04 0.21 Sources: U.S. DOT, BTS; U.S. DOE, EIA; AKRR; company proprietary information; author calculations.

Table 41. Comparison of estimated passenger transportation modes passenger-miles per gallon and fuel costs and emissions per passenger-mile transported, Alaska 2007–2010, 2011$

Railroad Ferry Air* Average passengers-miles per gallon of fuel 120 12 50 Average fuel costs per passengers-mile $0.02 $3.83 $0.09

CO2 emissions per passenger-mile (kilograms) 0.08 0.85 0.19

CO2/passenger-mile w/aviation multiplier 0.08 0.85 0.56 *U.S. average for comparison purposes only. Show with and without aviation emissions multiplier. Sources: U.S. DOT, BTS; U.S. DOE, EIA; U.S. Army Corps of Engineers, Waterborne Statistics; AMHS; IFA; Jardine, 2009; Ingram, 2008; company proprietary information; author calculations

Carbon Emissions Tax Analysis This section analyzes the potential impact of a carbon tax on Alaska industries and the air, water, rail and trucking transportation sectors. By using IMPLAN, we capture not only the direct fuel use of an industry but also the embedded energy intensity of all industry inputs. We do not analyze a specific tax proposal, since none is currently pending. Instead, we show the relative impact of a tax on carbon

emissions on the four transportation sectors analyzed. (U.S. EPA, 2006; Schipper, Unander, Murtishaw, Ting, 2001)

A carbon tax structured to decrease carbon emissions usually works by providing incentives for industries and households to adjust behavior by substituting away from the most carbon intensive products. This shift occurs because the prices of these products rise proportionately more than less- intensive ones. The carbon intensity of any given product is largely driven by the types of fossil fuels used in its production and by suppliers (Table 42).

Table 42. Industry sectors most impacted by a potential carbon emissions tax

Energy IMPLAN CO intensity "Tax" 2 Rank IMPLAN sector name Sector # Per unit of output 1 Petroleum refineries 115 0.1039 0.0104 2 Natural gas distribution 32 0.0698 0.0070 3 State and local government electric use 431 0.0452 0.0045 4 Asphalt paving mixture and block manuf. 116 0.0398 0.0040 5 State and local government passenger 430 0.0391 0.0039 6 Other basic chemical manufacturing 126 0.0370 0.0037 7 Transport by pipeline 337 0.0268 0.0027 8 Plastics material and resin manufacturing 127 0.0265 0.0026 9 Commercial fishing 17 0.0240 0.0024 10 Transport by air 332 0.0232 0.0023 11 Fertilizer manufacturing 130 0.0173 0.0017 12 Services to buildings and dwellings 388 0.0167 0.0017 13 Electric power generation transmission 31 0.0154 0.0015 14 Other state and local government enter. 432 0.0139 0.0014 15 Transport by truck 335 0.0120 0.0012 16 All other crop farming 10 0.0115 0.0011 17 All other chemical product and prep. 141 0.0111 0.0011 18 Toilet preparation manufacturing 139 0.0099 0.0010 19 Couriers and messengers 339 0.0094 0.0009 20 Extraction of oil and natural gas 20 0.0091 0.0009 21 Transit and ground passenger transp. 336 0.0083 0.0008 22 Cattle ranching and farming 11 0.0072 0.0007 23 Transport by rail 333 0.0068 0.0007 Source: IMPLAN Input-Output Model, U.S, DOE, EIA, author calculations.

To analyze the impact of a carbon dioxide tax on the transportation sectors of the Alaska economy, we used energy use estimates from the U.S. Department of Energy, Energy Information Administration, State Energy Data System (U.S. DOE, EIA, SEDS, 2012a) along with IMPLAN data on types of fuel and fuel usage intensity. The respective “tax rate” is based on the relative carbon dioxide intensity of the industry or service.

From the EIA, SEDS information, we obtained the statewide fuel consumption in Btus for diesel fuel, jet fuel, and residual fuel oil (U.S. DOE, EIA, 2012a). We used the IMPLAN statewide output to generate the Btus per dollar of output from which we generate industry/fossil fuel specific Btus per dollar of output. This result becomes the matrix of energy requirements in Btus for the transportation sectors based on their use of fossil fuels; this matrix provides an estimate of the energy requirements per unit of output/demand. This method is more comprehensive than simply applying a carbon tax to direct fuel usage, as it also includes energy used in production (Creedy and Sleeman, 2004a, b, 2005).

Table 43. Transportation sector fuel use and emissions by industry output

Output mmBtu/$ (mmBtu/$) * kg CO2/ "Tax" Tax w/air mmBtu (millions $$) of output emissions factor per unit of output multiplier Air 25,247,402 $1,817.4 0.014 0.985 1.541 0.0015 0.0045

Rail 471,216 $112.0 0.004 0.308 0.445 0.0004 0.0004

Truck 2,728,583 $569.0 0.005 0.351 0.488 0.0005 0.0005

Water 11,118,997 $541.7 0.021 1.617 2.491 0.0025 0.0025

Source: IMPLAN Input-Output Model, U.S, DOE, EIA, author calculations.

The calculated “tax” is the relative “tax” increase per unit of output, based on the carbon intensity of a unit of production of output. The highest kilogram of CO2 per unit of output is for water transportation followed by air, truck, and rail. The relatively higher carbon intensity of water transportation per unit of output as compared with air transportation per unit of output results primarily from the comparative advantage of air transportation’s higher value output per energy input (see column headed “mmBtu/$ of output” in Table 43). However, this initial result does not include an aviation emissions multiplier. As just discussed, these multipliers range from 1.9 to 4 for each kilogram of carbon emissions (Jardine, 2009). If an average multiplier of 2.9 is applied to the air carbon intensity per unit of output, the relative “tax” on air transportation is 0.0045, or 1.8 times greater than the comparative tax on water transportation. Therefore, a carbon tax would have a relatively higher impact per unit of output for air transportation than for water transportation following the imposition of an air emissions multiplier.

Given that the different transport sectors face different taxes, the consuming sectors would also face a varying additional burden, depending upon how much of each of these services they consume. It is noteworthy that while some sectors may not be large consumers of transport relative to the overall scale of the economy, the expenditures geared toward transport are a significant portion of their outlay. Therefore, the transport usage intensity—not just the dollar outlay—should be taken into account when considering such measures.

Energy used in production calculation method:

The intensity is defined by , which measures the kilograms of carbon dioxide emissions per final consumption of the output from industry (i). Therefore a carbon dioxide tax of α, which is placed on carbon dioxide emissions, is equivalent to an ad valorem tax exclusive tax rate on the commodity 1 group , where

As the intensity is expressed in terms of each dollar’s worth of the output that contributes to final demands, the total amount of carbon dioxide arising from all industries, E, is given by

∑ = y where is the value of final demand for industry i for i =1, n. The terms c and y denote corresponding column vectors, and the prime indicates transposition. The carbon dioxide intensities depend in a direct way on the types and amounts of fossil fuels used by each industry and the emissions per unit of those fossil fuels. However, the problem is complicated by the need to consider the total output of each industry, rather than merely the amount of that output which is consumed, that is the final demand.

Alaska fuel usage, expenditures, and intensity:

Calculations and variables: e' the k-element vector of CO2 emissions (kg of carbon dioxide) per unit of energy (Btu) associated with each of the k fossil fuels.

k fuels: Diesel fuel, Jet Fuel, Residual

F' n*k matrix matrix of energy requirements in Btus for n industries (transport) across k fuel types

n number of industries (transport)

k fuels

** Multiplying the transpose of the e vector by the transpose of the F matrix gives a row vector, which contains the carbon dioxide emissions per unit of gross output from each industry:

= ( ) Carbon dioxide intensities where ) is the total requirement matrix.

E Total carbon dioxide emissions (not comprehensive but as defined by the k fuels

1 For further exposition, see Creedy, J., and Sleeman, C. (2004b) Carbon Dioxide Emissions Reductions in New Zealand: A Minimum Disruption Approach. New Zealand Treasury and Creedy, J., and Sleeman, C. (2004a) Carbon Taxation, Prices and Household Welfare in New Zealand.

( )*Y

Y: final demand

α: ad hoc tax in this case 10$ per kg on carbon dioxide emissions

Tau(i) : α which refers to the ad valorem tax exclusive rate on the ith commodity group tau i.

From the EIA, SEDS information, we obtained the statewide fuel consumption in Btus for diesel fuel, jet fuel, and residual fuel oil. We used the IMPLAN statewide output to generate the Btus per dollar of output from which we generate an industry/fossil fuel-specific Btus per dollar of output. This calculation becomes our matrix of energy requirements in Btus for n transport industries across k fossil fuels, which provides an estimate of the energy requirements per unit of output/demand of the Alaska economy. This method is more comprehensive than simply applying a carbon tax to direct fuel usage.

Consider increasing the final consumption of a good by $1. The problem is to evaluate how much carbon dioxide this would involve. This increase in the final demand by $1 involves a larger increase in the gross, or total, output of the good, as well as requiring increases in the outputs of other goods, because intermediate goods, including the particular good of interest, are needed in the production process. The extent to which there is an increase in carbon dioxide depends also on the intermediate requirements of all goods, which are themselves intermediate requirements for the particular good. Indeed, the sequence of intermediate requirements continues until it “works itself out,” such that the additional amounts needed become negligible. This is in fact a standard multiplier process.

= ( )

This expression can then be used together with a selected carbon tax rate to calculate the effective carbon tax rates given by the following equation:

This analysis uses the industry input matrix of the IMPLAN software to account for the use of energy in the Alaska economy. In this way, we estimate both the fuel use and energy intensity per unit of output of final demand in the Alaska economy. Naturally, the most transport-intensive commodities are the ones that face the highest “tax” increase.

Conclusions In this analysis, we estimated energy and fuel use by Alaska transportation sectors to understand impacts of sudden increases in fuel price or an emissions tax on the transportation sectors and the Alaska economy. Results of our analysis indicate that Alaska major industries are most vulnerable to fuel price shocks. Alaska households are impacted directly and indirectly because most goods are shipped long distances to Alaska. We found that the inability to collect accurate data seriously hampered our ability to conduct this analysis. Even publically available data such as the BTS datasets are not

constructed or cross-referenced in a way that permits comprehensive regional fuel use, energy, and emissions analyses.

We estimated the energy and fuel used by the air, water, trucking, and rail transportation sectors in Alaska. We compared their fuel intensity to move passengers and freight by estimating their passenger- miles per gallon of fuel, ton-miles per gallon of fuel, and fuel costs per passenger-mile and per ton-mile.

We estimated that rail is the most efficient form of transportation for moving freight per gallon of fuel, followed by barge, marine ship, truck, and ferry. In estimating passenger-miles per gallon of fuel, we found again that rail transport is the most fuel-efficient, followed by air and ferry. Fuel costs per ton- mile and passenger-mile followed the same pattern of efficiency.

Faced with continued high or increasing fuel prices or carbon legislation, the demand for Alaska Railroad transportation services could potentially increase with shifts away from trucking. However, because the distance from Anchorage to Fairbanks is relatively short, freight handling would have to be quite efficient and wages competitive to compete with the comparative efficiency of truck transportation, with its fewer freight intermodal transfers.

We also found that between 2008 and 2010, employment in Alaska in the transportation and warehousing industry sector declined by approximately 10%. Air transportation employment declined by 15%, with scheduled airfreight transportation employment declining by 42%. Truck transportation employment declined by 15%, and water transportation employment expanded by 30%.

To connect transportation costs to the Alaska economy, we examined a number of factors including an assessment of the most transportation-intensive industries. The industries that are most dependent on transportation services, and thus more sensitive to changes in transportation costs, are:

1. Seafood product preparation and packaging 2. Support activities for oil and gas operations 3. Transport by truck 4. Drilling of oil and gas wells 5. Construction of new nonresidential commercial and health care structures 6. Construction of new residential single-/multi-family housing 7. Electric power generation, transmission, distribution 8. Mining of gold, silver, and other metal ore 9. Food services and drinking places 10. Other state and local government enterprises

As a result, these industries are the most impacted by increases in fuel prices or other impacts that raise the cost of transportation services as an input in their production. Most of these are core industries in the Alaska economy, so any impacts to these industries would have broad consequences.

Similarly, we looked at the Alaska industries that would be most affected by carbon emissions legislation. These industries are:

1. Petroleum refineries 2. Natural gas distribution 3. State and local government electric use 4. Asphalt paving mixture and block manufacturing 5. State and local government passenger 6. Other basic chemical manufacturing 7. Transport by pipeline 8. Plastics material and resin manufacturing 9. Commercial fishing 10. Transport by air

These are also the industries that could have the most payback from increased efficiencies and reduced dependence on fossil fuels in terms of avoiding potential emission tax impacts.

Because of the fuel and carbon intensity of air transportation, there has been sustained improvement in air transportation efficiency, with increased load factors and increased fuel efficiency of airplanes. Of the four transportation sectors analyzed, air transportation is the most vulnerable to emissions legislation impacts.

Alaska households at all income levels are also vulnerable to increases in the price of transportation services as a result of fuel price increases or carbon emissions legislation. If Alaska households continued to purchase transportation services at the same level after fuel price increases similar to those that occurred between 2008 and 2010, these services would cost an additional $26.8 million; an estimated 73% of these cost increases would be paid by households earning over $50,000 annually because they are more able to absorb these higher costs. In all likelihood, however, households would reduce their expenditures on transportation services. Water and truck transportation services declined the most in our simulation, probably because the majority of routine purchases of goods by Alaska households are transported by water and truck.

In addition to the higher cost of transportation services resulting from increases in fuel prices, direct purchases of refined petroleum products by Alaska households would cost an additional $124.1 million, if households continued to purchase at the same level after the kinds of fuel price increases experienced between 2008 and 2010. Similar to transportation services price increases, an estimated 70% of the refined petroleum price increases would be absorbed by households with incomes of $50,000 or higher.

Our economic impact simulation did not include utilities, so price increases for space heating and electricity are not included in these estimates.

References for Report and Appendices ADLWD (Alaska Department of Labor and Workforce Development, Research and Analysis), 2012, Consumer Price Index, Accessed December 1, 2012, http://labor.alaska.gov/research/cpi/cpi.htm ADOT&PF (Alaska Department of Transportation and Public Facilities), 2011, Alaska Aviation System Plan. ADOT&PF (Alaska Department of Transportation and Public Facilities), 2008, Let’s Get Moving 2030: Alaska Statewide Long-Range Transportation Policy Plan, February. ADOT&PF (Alaska Department of Transportation and Public Facilities), AMHS (Alaska Marine Highway System), 2012, Network Map. Alaska Air Carriers Association, 2012, Alaska Aviation Facts. Accessed August 12, 2012, http://www.alaskaaircarriers.org/ Cadavoa, L. 2010, Anchorage Fuel Sellers Report, Ted Stevens Anchorage International Airport, Phone# 907-266-2413. Personal communication. Chase, A. Robert, 2004, Ties That Bind: The Enduring Economic Impact of Alaska on the Puget Sound Region, Huckwell/Wieman Associates, Commissioned by the Tacoma-Pierce County Chamber of Commerce and Greater Seattle Chamber of Commerce. Creedy, J. and Sleeman, C., 2004a, Carbon Taxation, Prices and Household Welfare in New Zealand, Treasury Working Paper Series 04/23, New Zealand Treasury. Creedy, J. and Sleeman, C., 2004b, Carbon Dioxide Emissions Reductions in New Zealand: A Minimum Disruption Approach, New Zealand Treasury. Creedy, J. and Sleeman, C., 2005, Carbon Dioxide Emissions Reductions in New Zealand: A Minimum Disruption Approach, Australian Economic Papers, 44(3): 199–220. CSX, Inc. Company profile, http://www.csx.com/ Fisheries Economics Data Program, Monthly Marine Fuel Prices, http://www.psmfc.org/efin/data/fuel.html Fried, Neal and Brynn Keith, 2005, Transportation: Alaska geography makes it an important sector, Alaska Economic Trends, January. Ghanta, Praveen, 2010, Fuel Efficiency: Modes of Transportation Ranked by MPG, http://truecostblog.com/2010/05/27/fuel-efficiency-modes-of-transportation-ranked-by-mpg/ Government Accountability Office (GAO), 2011, Surface Freight Transportation: A Comparison of the Costs of Road, Rail, and Waterways Freight Shipments That Are Not Passed on to Consumers (GAO-11- 134). Gray, John T. and J. Phillip Rowe, 1982, An Alaska Census of Transportation, Institute of Social and Economic Research, University of Alaska Anchorage. Holguín-Veras, José, Jeffrey Wojtowicz, Carlos González-Calderón,, Michael Lawrence, Jonathan Skolnik, Michael Brooks, Shanshan Zhang, Anne Strauss-Wieder and Lóri Tavasszy, 2013, Freight Data Cost Elements, National Cooperative Freight Research Program, Transportation Research Board, NCFRP Report 22. Horizon Shipping Lines, 2012, company website.

ICF International, 2009, Comparative Evaluation of Rail and Truck Fuel Efficiency on Competitive Corridors, prepared for U.S. Department of Transportation, Federal Railroad Administration, Office of Policy and Communications. Ingram, George, 2008, Coping With Sky-High Jet Fuel Prices, Director, Airline Operations, Air Transport Association of America. Inter-Island Ferry Authority, 2012, http://www.interislandferry.com/ Jardine, Christian N., 2009, Calculating the Carbon Dioxide Emissions of Flights, Environmental Change Institute, Oxford University Centre for the Environment. Lee, Joosung J., Stephen P. Lukachko, Ian A.Waitz, and Andreas Schafer, 2001, Historical and Future Trends in Aircraft Performance, Cost, and Emissions, Annual Review Energy and Environment, 26: 167– 200. Leontief, W., 1936, Quantitative input and output relations in the economic system of the United States. Review of Economic Statistics, 18: 105–125. Leontief, W., 1966, Input-output economics. New York: Oxford University Press, 257 pp. Metz, Paul, Mark Taylor, Tom Brigham, Stephane Larocqu, Jana Pierce, Ashleigh Arledge, Jim Calvin, Erin Harrington, Scott Miller, Bob Lingwood, Dave Marshall, and Teresa Watts, 2011, Alaska Marine Highway System Analysis, prepared by the Alaska University Transportation Center for the Alaska Marine Highway System, Alaska Department of Transportation and Public Facilities. MIG, Inc., 2011, IMPLAN System (data and software), 502 2nd Street, Suite 301, Hudson, WI 54016 www.implan.com Northern Economics, Inc., Economic and Community Contributions of Selected Alaska Airports: 12 Case Studies, Prepared for Alaska Department of Transportation and Public Facilities, June 2011. Northern Economics, Inc., The Economic Contribution of the Aviation Industry to the Economy of Alaska, Prepared for Alaska Department of Transportation and Public Facilities as part of the Alaska Aviation System Plan, March 2009. Northern Economics, Inc., 2006, Port of Anchorage Consolidation and Distribution Concept Feasibility Study, prepared for the Anchorage Economic Development Corporation, April 2006. Oil Price Information Services, weekly and monthly refinery prices wholesale price data, multiple years, www.opisnet.com Peeters, P.M., J. Middel, and A. Hoolhorst, 2005, Fuel efficiency of commercial aircraft: An overview of historical and future trends, National Aerospace Laboratory, NLR-CR-2005-669. QGI Consulting, Alaska Canada Rail Link, 2006, Traffic Data Development for Regional Re-supply, Alaska, Work Package: A-1 (A), March 4, 2006. Schultz, Caroline, 2012, The Span of Alaska’s Railroads, Alaska Economic Trends, Alaska Department of Labor and Workforce Development, Research and Analysis Section, Volume 32, Number 3, March. Schipper, Lee, Fridtjof Unander, Scott Murtishaw, and Mike Ting, 2001, Indicators of Energy Use and Carbon Emissions: Explaining the Energy Economy Link, Annual Review Energy and Environment, 26: 49– 81. Siebe, 2012, Senior Airport Engineer, HDR, Alaska, personal communication, December 2012.

Stolzer, Alan J., 2002, Fuel Consumption Modeling of a Transport Category Aircraft Using Flight Operations Quality Assurance Data: A Literature Review, Journal of Air Transportation, 7(1). Tuck, Brad and Mary Killorin, 2004, Economic Significance of Alaska Railroad, prepared for Alaska Railroad Corporation, Institute of Social and Economic Research, University of Alaska Anchorage. The Center for Neighborhood Technology, Travel Matters, 2013, prepared under contract with the Transit Cooperative Research Program, sponsored by the Federal Transit Administration, in cooperation with the Transportation Research Board and the American Public Transportation Association. http://www.travelmatters.org/calculator/individual/methodology T.O.T.E., 2012, company website. TRB (Transportation Research Board), 2012a, Climate Change and Transportation: Summary of Key Information, Transportation Research Circular E-C164, July. TRB (Transportation Research Board), 2012b, Adapting to Climate Change: Building a Network of Solutions, TR News, July–August 2012, Number 281. UAA (University of Alaska Anchorage, College of Business & Public Policy, Department of Logistics and Port of Anchorage, Municipality of Anchorage), 2011, Alaska’s Lifeline: Cargo Distribution Patterns from the Port of Anchorage to Southcentral, Northern, Western and Southeast Alaska, February, U.S. Army Corps of Engineers, 2012a, Waterborne Commerce Statistics, Part 4: Pacific Coast, Alaska and Hawaii, 2006-2010, updated June 6, 2012, http://www.ndc.iwr.usace.army.mil/wcsc/wcsc.htm#2010 Waterborne Commerce of the United States (WCUS) and http://www.ndc.iwr.usace.army.mil/wcsc/pdf/wcuspac10.pdf U.S. Army Corps of Engineers, 2012b, Waterborne Transportation Lines of the Unites States 2010: Volume 3 – Vessel Characteristics, http://www.ndc.iwr.usace.army.mil/veslchar/pdf/wtlusvl3_10.pdf U.S. Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), 2012, Distances between United States Ports, 2012 (12th) edition., http://www.nauticalcharts.noaa.gov/nsd/distances-ports/distances.pdf U.S. DOE, EIA (U.S. Department of Energy, Energy Information Administration), 2012a, State Energy Data System, June 29, 2012. http://www.eia.gov/state/seds/hf.jsp?incfile=sep_sum/html/rank_use_per_cap.html U.S. DOE, EIA (U.S. Department of Energy, Energy Information Administration), 2012b, Voluntary Reporting of Greenhouse Gases Program, Fuel Emission Coefficients, Table 2: Carbon Dioxide Emission Factors for Transportation Fuels, http://www.eia.gov/oiaf/1605/coefficients.html#tbl2 U.S. DOE, EIA (U.S. Department of Energy, Energy Information Administration), 2007, Documentation for Emissions of Greenhouse Gases in the U.S. 2005, DOE/EIA-0638 (2005), October 2007, Tables 6-1, 6-4, and 6-5. U.S. Department of Transportation, Federal Aviation Administration, 2005, Advisory Circular Aircraft Weight and Balance Control, AC 120-27E, 6/10/05. U.S. DOT, RITA, BTS (U.S. Department of Transportation, Research and Innovative Technology Administration, Bureau of Transportation Statistics),2010, Aviation Library: http://www.transtats.bts.gov/databases.asp?Mode_ID=1&Mode_Desc=Aviation&Subject_ID2=0 U.S. DOT, RITA, BTS (U.S. Department of Transportation, Research and Innovative Technology Administration, Bureau of Transportation Statistics), Aviation data (BTS). 2012a. Air Carriers: T-100

Segment (All Carriers). Accessed October 21, 2012, http://www.transtats.bts.gov/DL_SelectFields.asp?Table_ID=293&DB_Short_Name=Air%20Carriers U.S. DOT, RITA, BTS (U.S. Department of Transportation, Research and Innovative Technology Administration, Bureau of Transportation Statistics), Aviation data (BTS). 2012b. Air Carrier Traffic and Capacity Statistics by Aircraft Type: Schedule T2. Accessed October 21, 2012, http://www.transtats.bts.gov/Fields.asp?Table_ID=254 U.S. DOT, RITA, BTS (U.S. Department of Transportation, Research and Innovative Technology Administration, Bureau of Transportation Statistics), Aviation data (BTS). 2012c. Air Carrier Financials: Schedule P-12 (a). Accessed October 21, 2012, http://www.transtats.bts.gov/DL_SelectFields.asp?Table_ID=294&DB_Short_Name=Air%20Carrier%20Fi nancial U.S. DOT, RITA, BTS (U.S. Department of Transportation, Research and Innovative Technology Administration, Bureau of Transportation Statistics), Aviation data (BTS). 2012d. Form 41, Schedule T100 and T100(f) Air Carrier Data. Accessed on October 21, 2012, http://www.bts.gov/programs/statistical_policy_and_research/source_and_accuracy_compendium/for m41_schedule.html U.S. Department of Transportation, Federal Highway Administration, Highway Statistics 2009, 2010, Table VM-1, April 2010. http://www.fhwa.dot.gov/policyinformation/statistics/2009/vm1.cfm U.S. Environmental Protection Agency, 2008, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 to 2006. Verrelli, Sarah, 2012, Work Place Deaths on Steady Decline, Alaska Economic Trends, Alaska Department of Labor and Workforce Development, Research and Analysis Section, Volume 32, Number 3, March. Wells, Douglas P., 2011, NASA Green Flight Challenge: Conceptual Design Approaches and Technologies to Enable 200 Passenger Miles per Gallon, NASA Langley Research Center. Winebrake, James J. and James J. Corbett, Improving the Energy Efficiency and Environmental Performance of Goods Movement, in Climate and Transportation Solutions: Findings from the 2009 Asilomar Conference on Transportation and Energy Policy, edited by Daniel Sperling and James S. Cannon (Institute of Transportation Studies, University of California, Davis, 2010).

Appendix A. Marine Transportation Companies Marine Ships Totem Ocean Trailer Express (TOTE, Inc.) is an Alaska-based transportation company offering marine and land transportation services between Alaska and the contiguous U.S. states. TOTE’s service offering includes both marine and highway services operating between Tacoma, Washington and Anchorage, Alaska (TOTE, 2012).

TOTE operates a twice-weekly northbound and southbound service between Tacoma and Anchorage. The marine service is a roll on-roll off service for highway trailers and automobiles. The ORCA and Ponce class vessels have trailer capacities of 600 and 380 forty-foot equivalent units (FEU), respectively. Transit time between Tacoma and Anchorage ranges from 66 to 72 hours in each direction over a sailing distance of 1,450 nautical miles.

Horizon, Inc., based in Charlotte, North Carolina, is the United States’ leading Jones Act container shipping and logistics company, accounting for approximately 37% of the total U.S. marine container shipments between the continental U.S. and the three noncontiguous Jones Act markets of Alaska, Hawaii, and Puerto Rico. In 2003 the Carlyle Group, a global private equity firm, purchased the Horizon Services Group of CSX Lines and renamed it Horizon Lines LLC.

Horizon operates twice-weekly service from Tacoma directly to Anchorage with a follow-on call to the Port of Kodiak. One vessel per week then makes a call at Dutch Harbor. Truck and barge services connect these three principal destination ports with surrounding locations including Akutan, Bristol Bay, the Pribilof Islands, King Cove, Sand Point, the Kenai Peninsula, Prudhoe Bay, Eagle River, Fairbanks and Palmer. Transit time between Tacoma and Anchorage is approximately three and one-half days (Horizon, 2012).

Barges Lynden Inc. is the parent company of a family of transportation and logistics companies primarily serving Alaska and the Pacific Northwest. Lynden companies provide multi-modal transportation services including air, marine, and land services to, from, and within Alaska. Key transportation companies under the Lynden Group include:

Alaska Marine Lines Alaska Marine Lines (AML) provides twice per week barge services between Seattle and Southeast Alaska. The Southeast Alaska barge services a number of locations including Juneau, Ketchikan, Petersburg, Sitka, and Kake. AML barge services handle full container, less than container, refrigerated, and break bulk cargo. Principal customers consist of retail establishments such as grocery outlets.

AML also services the Central Alaska market with weekly barge service between Seattle and Anchorage with overland connections to Seward, Kenai, and Fairbanks.

-A-1-

Alaska West Express Alaska West Express (AWE) provides truckload transportation throughout the United States and Canada, specializing in shipments to and from Alaska. Alaska West Express is a leader in transporting liquid and dry-bulk products, hazardous and nonhazardous chemicals and petroleum products. AWE operates a rail terminal at Fairbanks, offering product transfer services for liquid and dry bulk products serving primarily the oil and gas industry.

Lynden Transport Lynden Transport is a complete multi-modal, regional, common and contract carrier primarily serving Alaska. Lynden Transport also provides LTL cargo service on motor-water-motor routes using steamships, barges, and ferries. Lynden Transport has truckload capabilities for dry van, refrigerated, flatbed and heavy-haul commodities on both water and highway routes.

Alaska Railbelt Marine (ARM) operates scheduled, once per week railcar barge service between Seattle and Whittier, Alaska. This service operates in partnership with the Alaska Railroad providing freight service to Southcentral Alaska. ARM operates the service using three rail barges with estimated railcar capacities of 40 cars each plus deck space for bulk and container cargo.

CN AquaTrain CN AquaTrain, operated by Foss Maritime, provides marine services for railcar movements between Alaska and Canada and the Lower 48.2 The service operates one of the world's largest railcar barges, accommodating 45 railcars on 8 tracks. The service is integrated with CN Rail’s North American network and services shippers of bulk products serving the Alaska market principally with industrial commodities such as fertilizers, sand, methanol, salt, lumber, and petroleum products. The CN AquaTrain service connects to the Alaska Railroad at Whittier for rail delivery to inland points. Transit time over the 830 nautical mile voyage is approximately four days.

CN AquaTrain operates an average of 35 barge sailings per year from Prince Rupert, British Columbia, to Whittier, Alaska. With a nominal barge capacity of 45 railcars per sailing, its estimated total annual movements are in the order of 1,575 railcars per year. Based on an average per car payload of 92 tons per railcar, we estimate that CN’s AquaTrain service can transport approximately 145,500 short tons of freight into the Alaska market annually.

2 Foss Maritime was purchased by Saltchuk Resources in 1987. Saltchuk is also the parent company of Totem Ocean Trailer Express.

-A-2-

Appendix B. Barge Fuel Use Calculations In contrast to other modes where we received considerable information from a number of companies or the data were public through reporting requirements, only one barge company provided monthly data for the movement of freight. We used this barge company’s data as a prototype to construct a barge fuel use model. The prototype barge company’s information was expanded to include additional barge trips by travel segments based on U.S. Army Corps of Engineers published Waterborne Statistics of freight movement by port (U.S. Army Corps of Engineers, 2012a, b) and published freight schedules of barge companies serving Alaska (company websites, Northern Economics, 2006).

Scheduled Barge Traffic For scheduled barge service, we used the number of trips posted on company websites and other sources to calculate an estimated number of barge miles to the Southcentral, Southeast, and Western regions of Alaska as described below.

Southcentral Alaska For this calculation, we used number of trips between Anchorage-Seattle and Whittier-Seattle. This number of trips was multiplied by 3,000, the Anchorage-Seattle / Whittier-Seattle distance traveled provided by the prototype barge company. Some of the trips between Anchorage-Seattle also include additional travel to Western Alaska, so to avoid double counting, 1,500 miles was subtracted for every trip to Western Alaska that stopped in Anchorage. This was the same methodology used for trips to Western Alaska that the prototype barge company provided in their calculations.

Nautical Miles to Southcentral Alaska = (w * 3,000) + (a * 3,000) – (W *1,500) w = # of trips to Whittier a = # of trips to Anchorage

W = # of trips to Western Alaska

Southeast Alaska For the calculation of the number of miles traveled by the prototype barge company to Southeast Alaska, we started with number of trips then multiplied by 1,770, the provided number of miles. This was adjusted, as appropriate, for any trips that are shared with other barge companies.

Nautical Miles to Southeast Alaska = (SE * 1,770)

SE = # of trips to Southeast Alaska

Western Alaska Western Alaska trips reflected the seasonal closure by Bering Sea ice and incorporated the other portion of miles from a trip with an Anchorage link. For the months of October to February, we multiplied the number of trips per month by 5,000, in March by 6,000, and during the months of April to September by 6,800, reflecting the data provided by the prototype barge company.

-B-1-

If from October through February Nautical Miles to Western Alaska = (W * 5,000) – (a *1,500) If in March Nautical Miles to Western Alaska = (W * 6,000) – (a *1,500) If from April through September Nautical Miles to Western Alaska = (W * 6,800) – (a *1,500) a = # of trips to Anchorage

W = # of trips to Western Alaska

Unscheduled Barge Traffic

Kivalina The Port of Kivalina is unique in Western Alaska because of the nearby Red Dog mine, which is included in the Kivalina data in U.S. Army Corps of Engineers, Waterborne Statistics port information. For example, nearly three million short tons were exported in 2010, making it (by tonnage) the second largest exporting port in Alaska.3 Due to the shallow nature of the port, numerous trips are made to lighter materials by barge to approximately 5 miles offshore to transfer ore to larger cargo ships and to transport fuel needed for the mine’s operation. Another characteristic of Kivalina is the short window that vessels can navigate the Bering Sea, giving roughly four months to perform all this needed shipping.

To calculate the nautical miles traveled by tug to service Kivalina, we used the total number of barges that left the port and multiplied this by the estimated 10 miles traveled to lighter to the barge and return. Due to seasonal ice constraints, it is assumed that efforts to transfer materials would be evenly distributed across the months of June through September.

Liquid Barge Traffic Several assumptions were needed to calculate liquid barge traffic in Alaska. Since we had data from waterborne statistics on the number of vessels that departed and received at a port but no way to identify the exact route, we assumed that the supply chain for Alaska was based on a three-stage system. An example of this process is as follows: first stage we assumed was refinery to hub, second stage was hub to sub-hub, and the third stage was sub-hub to communities. The third stage is the least- observable stage because of the limits of the available public data and the proprietary nature of the data. If one of these stages was skipped because it was more economical to service directly to a community, that would not be observable for the same reason that the third stage is not observable.

3 U.S. Army Corps of Engineers, Waterborne Commerce Statistics, Pacific Coast, Alaska, 2012.

-B-2-

To calculate the distances traveled, we used the NOAA distance table4 when possible or extrapolated distance combining the NOAA distance and the distance provided by the prototype barge company (NOAA, 2012).

Valdez – Anchorage As a first stage in the supply chain, we assumed fuel from the refinery in Valdez was shipped directly to Anchorage. For the quantity of trips in this route we used the total number of liquid barges that departed from Valdez and assumed each barge was an individual trip with a single tug with each barge. For distance, we used the distance provided by the National Oceanic and Atmospheric Association (NOAA) in its published port-to-port distances (NOAA, 2012). The distribution of trips is again based on the prototype barge company’s seasonal distribution for Southcentral Alaska.

Anchorage – Unalaska This is an example of a second stage, where Unalaska is seen as a sub-hub to Western Alaska. The quantity of liquid barges received at Unalaska is assumed to be each a single trip between Unalaska and Anchorage, and the distance is based on the data provided by the prototype barge company.

Unalaska – Western Alaska Because of the lack of information about supplying in this region, we used the distance provided by the prototype barge company data and the number of trips from waterborne statistics. For the number of trips, we used the number of liquid barges that left Unalaska, and for the schedule, we used the seasonal ice-free distribution of trips for Western Alaska by the prototype barge company. For the distance, we used the difference between 5,000 (distance from Dutch Harbor to Seattle) and 6,800 (assumed distance from Seattle to Western Alaska). This resulted in a calculated 1,800 nautical miles per trip to supply liquid fuels to the coast of Western Alaska from Unalaska/Dutch Harbor. These calculations are meant to capture both the second and third stage in our supply chain model.

Anacortes, Washington – Southeast Alaska For this calculation, we used the distance the prototype barge company provided to service Southeast Alaska and subtracted the distance between Tacoma, Washington, and Anacortes, Washington. For the number of trips, we used the number of liquid barges making deliveries to Ketchikan, the major fuel hub for this region, shown in Waterborne Statistics. We assume that the additional trips needed to service the other communities in the region are captured by the miles provided by the prototype company.

Final Calculations

Marine Service After constructing the model of the number of nautical miles traveled in transporting barges by tug, we then had to calculate fuel usage for these trips. For these calculations we used the figures provided by the prototype barge company for daily fuel usage by region. They also used the average speed of 7.5 nautical miles per hour in their calculations. Given the lack of information provided by other barge

4 Distance Between United States Ports - http://www.nauticalcharts.noaa.gov/nsd/distances-ports/distances.pdf

-B-3-

companies, we had to assume that the prototype barge company’s tugs were sufficiently similar in operation and condition.

The prototype barge company gave a different fuel usage per day for each region. For Southeast Alaska they provided 3,000 gallons per day, for Southcentral Alaska they provided 3,800 gallons per day, and for Western Alaska they provided 4,500 gallons per day. We believe these figures incorporate the additional fuel usage for tugs having to idle for proper tide levels to offload, varying ocean conditions, and the variation in the fleet for each region.

With the average of 7.5 nautical miles per hour, we calculated the number of hours and then the number of days traveled per month to service each region. Taking the number of days traveled by region, we used the daily fuel usage figures provided by the prototype barge company to calculate the total fuel usage in each region.

Inland Waterways The process we followed for estimating inland barge traffic was based on a combination of Waterborne Statistics, U.S. Census Data, and expert input. The major rivers we were concerned with in these calculations were the Kuskokwim and the Yukon Rivers, given the significant populations along or near their banks and the amount of supply required to support these communities.

The calculations began with the 2010 U.S. Census Data, which was then combined with the approximation that 1.8 to 2.7 tons of petroleum is required per person in Western Alaska (Northern Economics, 2006). With these two figures, we calculated the petroleum needed for each community along the Yukon and Kuskokwim Rivers. We used the upper end of the per capita estimate in part to also capture the additional transportation for other supplies.

Given the heterogeneity of both the vessels that operated along these waterways and the conditions along these waterways, we had to make simplifying assumptions for our calculations.

An important assumption we made was for a prototype tug and barge set. The purpose of this was to combine observation of ships listed on inland waterway ports5 and expert information into a simplified form for making equal calculations across these waterways. Our prototype tug and barge set consisted of a tug with 1,500 horsepower transporting four barges capable of safely transporting 120,000 gallons6 or roughly 1,720 short tons per sailing. Another assumption made for operating conditions was an average operating speed of four knots, and average operational level at 66% of maximum horsepower. Under these assumptions we calculated fuel usage based on 40 gallons per 1,000 horsepower per hour. We were informed that barge service along the river systems operates under the process of supplying each community as the barge reaches it; the barge then proceeds along the river until it is empty. Using the prototype tug and barge set with the expert information, we used two modified methods for calculating transportation on each of the river systems.7

5 U.S. Army Corps of Engineers, Waterborne Transportation Lines of the Unites States 2010 6 Conversion: 1 short ton ~ 279 gallons, Source: U.S. Geological Survey Digital Data Series 7 Several of these assumptions are based on information provided by Mark Smith, Vitus Marine

-B-4-

Service to communities along the Yukon River originates from one of two points: Communities upstream of Pilot Station are serviced from the city of Nenana; all other communities including Pilot Station are serviced through the mouth from a source along the ocean.8 But because the Nenana River joins the Yukon River at Tanana and there is a significant population upstream from this junction, we separated these communities into two groups, with the river junction being the dividing point.9 With the communities sectioned into three categories, the number of trips needed to service them was calculated based on the assumed tug and barge set described above. The distance to the communities was found by combining distances from NOAA,10 with estimates found by string measurement using Google Earth.

The calculations for servicing communities along the Kuskokwim River were very similar to those described in servicing the Yukon River. We started by dividing the river into three sections. Since services for these communities are based largely out of the city of Bethel, the communities were divided based on their distance from Bethel. The first section was for all those communities within 50 nautical miles of Bethel. With the population of Bethel, these communities make up nearly 85% of the population along the Kuskokwim River. The second section was those communities between 50 to 100 nautical miles, and the third section was all remaining communities greater than 100 nautical miles from Bethel. We then calculated the number of trips based on the total trips needed from the prototype tug and barge set to service each section. For total distance traveled we multiplied the maximum distance traveled to service a section by the total trips found. The distances for these calculations were all based on string measurement using mapping software.

Fuel usage = (D/ 4 knots) * (HP) *((40 Gallons/hour)/ 1000 HP)

D = Distance traveled in Nautical Miles

HP = Average Operational Horsepower

Limitation of the Analysis

Lack of Inland Traffic Data Because of the lack of public data on barge traffic on inland waterways, we had to estimate based on a combination of assumptions, expert references, and calculations using inferred information.

Minor Traffic A concern that exists in all these estimates is the lack of data on third-stage distribution and any other minor barge traffic to specific communities. Many communities had no data available through U.S. Waterborne Statistics. Communities that did have data lacked unique identification, which made the risk of double counting an issue. In an effort to avoid double counting, we likely estimated too low rather than too high.

8 Based on information provided by Vitus Marine 9 All communities that are along the Yukon River but have road access were excluded because of the availability of trucks to supply them. 10 Distance Between United States Ports - http://www.nauticalcharts.noaa.gov/nsd/distances-ports/distances.pdf

-B-5-

Appendix C. Data Dictionary of Variables and Sources Used for Aviation Fuel Estimates Name Source Description FREIGHTenter BTS - T100 Freight in lbs on unscheduled flights coming into Alaska from outside Alaska

FREIGHTenter_s BTS - T100 Freight in lbs on scheduled flights coming into Alaska from outside Alaska

FREIGHTexit BTS - T100 Freight in lbs on NON-scheduled flights exiting Alaska

FREIGHTexit_s BTS - T100 Freight in lbs on scheduled flights exiting Alaska

FREIGHTintra BTS - T100 Freight in lbs on NON-scheduled intra-Alaska flights

FREIGHTintra_s BTS - T100 Freight in lbs on scheduled intra-Alaska flights

MAILenter BTS - T100 Mail in lbs on NON-scheduled flights coming into Alaska from outside Alaska

MAILenter_s BTS - T100 Mail in lbs on scheduled flights coming into Alaska from outside Alaska

MAILexit BTS - T100 Mail in lbs on NON-scheduled flights exiting Alaska

MAILexit_s BTS - T100 Mail in lbs on scheduled flights exiting Alaska

MAILintra BTS - T100 Mail in lbs on NON-scheduled intra-Alaska flights

MAILintra_s BTS - T100 Mail in lbs on scheduled intra-Alaska flights

MILESenter BTS - T100 Sum of miles in between all segments of NON-scheduled flights originating outside Alaska with a destination in Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Seattle and Anchorage, the distance shown in miles would be the distance between Seattle and Anchorage instead of three times the distance between Seattle and Anchorage.

MILESenter_s BTS - T100 Sum of miles in between all segments of scheduled flights originating outside Alaska with a destination in Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Seattle and Anchorage, the distance shown in miles would be the distance between Seattle and Anchorage instead of three times the distance between Seattle and Anchorage.

MILESexit BTS - T100 Sum of miles in between all segments of NON-scheduled flights exiting Alaska with a destination outside Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Anchorage and Seattle, the distance shown in

-C-1-

miles would be the distance between Seattle and Anchorage instead of three times the distance between Seattle and Anchorage.

MILESexit_s BTS - T100 Sum of miles in between all segments of scheduled flights exiting Alaska with a destination outside Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Anchorage and Seattle, the distance shown in miles would be the distance between Seattle and Anchorage instead of three times the distance between Seattle and Anchorage.

MILESintra BTS - T100 Sum of miles in between all segments of NON-scheduled flights within Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Fairbanks and Anchorage, the distance shown in miles would be the distance between Fairbanks and Anchorage instead of three times the distance between Fairbanks and Anchorage.

MILESintra_s BTS - T100 Sum of miles in between all segments of scheduled flights within Alaska. This is NOT equal to the distance flown, but rather the distance between all the segments. If a flight had 3 trips between Fairbanks and Anchorage, the distance shown in miles would be the distance between Fairbanks and Anchorage instead of three times the distance between Fairbanks and Anchorage.

PAXenter BTS - T100 Passengers on NON-scheduled flights coming into Alaska from outside Alaska

PAXenter_s BTS - T100 Passengers on scheduled flights coming into Alaska from outside Alaska

PAXexit BTS - T100 Passengers on NON-scheduled flights exiting Alaska

PAXexit_s BTS - T100 Passengers on scheduled flights exiting Alaska

PAXintra BTS - T100 Passengers on NON-scheduled intra-Alaska flights

PAXintra_s BTS - T100 Passengers on scheduled intra-Alaska flights

NALAcost BTS - P12A Nominal fuel cost of NON-scheduled intra-Alaska flights

NALAcost_real author calc. Inflation adjusted NALA_cost using the U.S. CPI as shown at: http://labor.alaska.gov/research/cpi/cpi.htm and converting to 2011 U.S. $

NALAgallons BTS - P12A Fuel consumption of NON-scheduled intra-Alaska flights

SALAcost BTS - P12A Nominal fuel cost of scheduled intra-Alaska flights

SALA_cost_real author calc. Inflation adjusted SALA_cost using the U.S. CPI as shown at: http://labor.alaska.gov/research/cpi/cpi.htm and converting to 2011 U.S. $

SALAgallons BTS - P12A Fuel consumption of scheduled intra-Alaska flights

-C-2-

FUEL EIA SEDS Transportation sector fuel consumption of aviation gas and jet fuel in gallons, equal to fuel consumption of intra-Alaska flights and flights exiting Alaska. Originally, these data were annual and av-gas and jet fuel were summed. I reallocated based on seasonality observed in intra-Alaska fuel consumed on scheduled and non-scheduled flights.

Note, through 2004, the EIA data includes kerosene-type (Jet A) and naphtha-type (Jet B) jet fuel. Beginning in 2005, data only include kerosene-type jet (A) fuel. Naphtha-type jet fuel, which is used by the military, is included in EIA SEDS "Industrial sector, Other Petroleum."

Note, fuel consumption by the military is not included in the BTS data.

FUELcost EIA Transportation sector fuel cost of aviation gas and jet fuel in gallons equal to fuel costs related to intra-Alaska flights and flights exiting Alaska. Just like the data on fuel (see above), these data are aggregated into annual numbers. Since we have much more granular data for aviation, it makes sense to use the best available data we have, so for intra-Alaska flights, we use the monthly fuel price information. Since the entering and exiting fuel quantities are estimates based on EIA SEDS estimates, it seems like it would be reasonable to use the annual fuel price figures that accompany that data. Otherwise, the effort required to produce precision in fuel quantity data may not be supported.

FUELcost_real author calc. The product of FUELcost and an annual average price calculated in sheet: EIA_SEDS_fuel_cost inflation adjusted using the U.S. CPI as shown at: http://labor.alaska.gov/research/cpi/cpi.htm and converting to 2011 U.S. $

CONFIG (Aircraft configuration) Code Description 0 Aircraft Configuration Not Relevant 1 Passenger Configuration 2 Freight Configuration 3 Combined Passenger and Freight on a main deck 4 Seaplane 9 Used for capturing expenses not attributed to specific aircraft types

CLASS (service class) Note, highlighted are the classes observed in the data Code Description A Scheduled First Class Passenger/ Cargo Service A C Scheduled Coach Passenger/ Cargo Service C E Scheduled Mixed First Class and Coach, Passenger/ Cargo Service E F Scheduled Passenger/ Cargo Service F G Scheduled All Cargo Service G H Humane Reason Unscheduled, Non-Revenue-Generating K Scheduled Service K (F+G) L Non-Scheduled Civilian Passenger/ Cargo Service L N Non-Scheduled Military Passenger/ Cargo Service N P Non-Scheduled Civilian All Cargo Service P Q Non-Scheduled Services (Other Than Charter) Q R Non-Scheduled Military All Cargo R

-C-3-

V Non-Scheduled Service V (L+N+P+R) For U.S. Carrier and (L+P+Q) For Foreign Carrier Z All Service Z (K+V)

-C-4-

Table C1. Fuel usage and costs for intra-state scheduled air transportation in Alaska 2005-2010, 2011$ Scheduled intra-state 2005 2006 2007 2008 2009 2010 Fuel data Total reported fuel consumption [gallons] 32,256,258 32,654,457 30,684,229 31,457,698 27,330,122 27,846,200 Total reported fuel cost [2011$] $70,049,491 $80,488,158 $79,832,197 $103,765,976 $55,388,981 $69,842,835 Reported average price [$/gal] $2.17 $2.46 $2.60 $3.30 $2.03 $2.51 T100 segment data 1 Passenger flights Passengers 2,268,370 2,252,414 2,508,783 2,419,234 2,129,342 2,007,724 Freight [short tons] 19,425 17,897 18,160 14,318 15,334 15,755 Mail [short tons] 31,569 29,925 32,897 32,516 32,562 33,967 Total weight [short tons] 277,831 273,063 301,935 288,757 260,830 250,495 Distance [statute miles] 30,471,202 29,325,912 31,514,108 31,054,248 28,970,104 27,262,634 Fuel [gallons] 17,469,905 17,368,112 17,461,721 17,146,127 14,954,123 14,032,350 Fuel cost[$2011] $37,938,620 $42,809,695 $45,430,751 $56,558,005 $30,306,985 $35,195,434 2 Cargo flights Passengers 1,279 1,153 83 - 9 - Freight [short tons] 61,067 60,524 64,613 63,674 54,787 57,504 Mail [short tons] 65,252 65,279 63,679 63,781 59,005 58,621 Total weight [short tons] 126,447 125,918 128,300 127,454 113,793 116,125 Distance [statute miles] 10,044,037 9,735,366 10,414,435 10,299,871 10,053,010 10,046,743 Fuel (gallons) 7,950,910 8,009,018 7,419,954 7,568,123 6,524,065 6,505,137 Fuel cost $17,266,640 $19,740,985 $19,304,746 $24,964,116 $13,222,089 $16,315,950 3 Mixed flights Passengers 656,059 728,181 607,596 700,017 625,298 905,776 Freight [short tons] 17,177 17,108 16,056 18,844 19,434 20,494 Mail [short tons] 19,873 19,189 16,590 17,187 13,497 13,050 Total weight [short tons] 102,656 109,116 93,406 106,032 95,461 124,121 Distance [statute miles] 11,994,249 12,525,908 10,512,191 10,158,530 7,908,228 10,145,191 Fuel (gallons) 6,454,955 6,940,278 5,401,903 6,296,086 5,473,034 6,953,071 Fuel cost $14,017,940 $17,106,705 $14,054,314 $20,768,191 $11,092,002 $17,439,442 4 Seaplane Passengers 42,824 34,509 49,308 56,515 47,347 44,837 Freight [short tons] 491 542 631 595 621 625 Mail [short tons] 1,278 1,307 1,366 1,287 1,253 1,239 Total weight [short tons] 6,051 5,299 6,928 7,534 6,609 6,349 Distance [statute miles] 824,276 789,645 923,078 983,336 968,210 987,847 Fuel (gallons) 380,489 337,049 400,650 447,362 378,901 355,642 Fuel cost $826,291 $830,773 $1,042,385 $1,475,664 $767,904 $892,009

Notes: Numbers in black are directly from the BTS and EIA data. Blue numbers were calculated from BTS and EIA data. Total weight includes freight, mail and passengers. Aircraft type numbers are T100 data codes. Sources: U.S. DOT, BTS; U.S. DOE, EIA; FAA; author calculations.

-C-5-

TableC2. Fuel usage and costs for intra-state non-scheduled air transportation in Alaska 2005-2010, 2011$

Non-scheduled intra-state 2005 2006 2007 2008 2009 2010 Fuel data Total reported fuel consumption [gallons] 1,463,625 1,388,093 5,963,125 9,301,368 6,732,250 7,859,375 Total reported fuel cost [2011$] $3,435,273 $3,587,177 $17,190,711 $32,761,806 $16,001,674 $17,258,454 Reported average price [$/gal] $2.35 $2.58 $2.88 $3.52 $2.38 $2.20 T100 segment data 1 Passenger flights Passengers 127,981 119,216 126,166 103,532 106,563 105,223 Freight [short tons] 1,794 1,559 1,937 1,214 1,174 1,283 Mail [short tons] 10 13 42 25 62 68 Total weight [short tons] 14,602 13,494 14,596 11,592 11,893 11,873 Distance [statute miles] 3,543,858 3,468,474 3,758,337 3,243,212 3,445,972 3,521,271 Fuel [gallons] 537,512 462,930 1,582,156 2,982,032 2,298,728 2,505,250 Fuel cost[$2011] $1,261,594 $1,196,325 $4,561,097 $10,503,484 $5,463,774 $5,501,295 2 Cargo flights Passengers 1 - 20 - - - Freight [short tons] 16,099 18,103 31,648 16,473 14,428 16,246 Mail [short tons] 518 317 74 56 142 23 Total weight [short tons] 16,617 18,420 31,724 16,529 14,570 16,269 Distance [statute miles] 1,154,873 892,210 901,922 561,937 472,310 590,498 Fuel (gallons) 611,704 631,938 3,438,809 4,251,946 2,816,141 3,432,777 Fuel cost $1,435,730 $1,633,084 $9,913,522 $14,976,444 $6,693,596 $7,538,058 3 Mixed flights Passengers 32,977 37,401 35,311 34,240 36,658 43,049 Freight [short tons] 1,692 1,798 1,492 1,251 1,184 1,101 Mail [short tons] 36 42 35 27 21 20 Total weight [short tons] 5,026 5,580 5,059 4,702 4,871 5,426 Distance [statute miles] 1,087,000 1,116,981 1,145,982 971,469 922,759 1,077,665 Fuel (gallons) 185,025 191,446 548,348 1,209,643 941,458 1,144,939 Fuel cost $434,272 $494,745 $1,580,796 $4,260,672 $2,237,723 $2,514,180 4 Seaplane Passengers 34,982 28,852 35,947 32,435 34,386 36,236 Freight [short tons] 16 81 38 87 56 53 Mail [short tons] 1 1 - 4 2 3 Total weight [short tons] 3,515 2,967 3,633 3,334 3,497 3,680 Distance [statute miles] 133,947 111,134 204,976 170,635 123,687 110,298 Fuel (gallons) 129,384 101,780 393,812 857,747 675,923 776,409 Fuel cost $303,676 $263,024 $1,135,295 $3,021,207 $1,606,580 $1,704,922

-C-6-

Table C3. Fuel usage and costs for exiting scheduled and non-scheduled air transportation in Alaska 2005-2010, 2011$

Exiting (scheduled & nonscheduled) 2005 2006 2007 2008 2009 2010 Fuel data Total reported fuel consumption [gallons] 1,319,394,117 1,309,831,450 1,193,994,646 967,954,934 762,383,628 925,632,425 Total reported fuel cost [2011$] $2,697,422,794 $2,994,219,343 $2,856,751,124 $3,063,036,030 $1,427,105,264 $2,169,344,617 Reported average price [$/gal] $2.04 $2.29 $2.39 $3.16 $1.87 $2.34 T100 segment data 1 Passenger flights Passengers 2,189,746 2,154,740 2,195,794 2,032,243 1,815,705 1,865,517 Freight [short tons] 22,171 17,123 17,166 11,891 13,605 13,663 Mail [short tons] 265 401 360 305 279 204 Total weight [short tons] 241,411 232,999 237,105 215,421 195,454 200,419 Distance [statute miles] 33,510,436 31,287,722 30,879,363 29,058,494 25,152,916 26,741,639 Fuel [gallons] 101,982,850 93,977,168 87,607,285 78,195,517 68,502,012 69,581,079 Fuel cost[$2011] $208,497,870 $214,827,835 $209,609,155 $247,445,081 $128,228,858 $163,072,656 2 Cargo flights Passengers 0 0 0 0 0 0 Freight [short tons] 2,870,685 3,007,132 2,986,227 2,442,106 1,972,402 2,457,062 Mail [short tons] 7,663 6,393 4,412 3,881 2,376 3,797 Total weight [short tons] 2,878,348 3,013,525 2,990,639 2,445,987 1,974,778 2,460,859 Distance [statute miles] 137,047,042 142,574,848 137,020,314 111,727,133 87,638,070 108,131,950 Fuel (gallons) 1,215,943,618 1,215,468,698 1,105,002,697 887,868,211 692,112,272 854,355,694 Fuel cost $2,485,924,402 $2,778,510,079 $2,643,829,021 $2,809,606,340 $1,295,564,372 $2,002,297,970 3 Mixed flights Passengers 20,966 6,053 20,136 30,563 28,996 27,652 Freight [short tons] 1,375 346 1,728 2,144 2,140 2,111 Mail [short tons] 2 5 6 10 9 8 Total weight [short tons] 3,474 956 3,748 5,210 5,048 4,884 Distance [statute miles] 399,219 107,917 387,368 614,689 648,313 676,972 Fuel (gallons) 1,467,649 385,584 1,384,665 1,891,206 1,769,344 1,695,652 Fuel cost $3,000,522 $881,429 $3,312,948 $5,984,609 $3,312,034 $3,973,990 4 Seaplane Passengers 0 0 0 0 0 0 Freight [short tons] ------Mail [short tons] ------Total weight [short tons] ------Distance [statute miles] ------Fuel (gallons) ------Fuel cost $0 $0 $0 $0 $0 $0

-C-7-

Table C4. Fuel usage and costs for entering scheduled and non-scheduled air transportation in Alaska 2005-2010, 2011$

Entering (scheduled & non-scheduled) 2005 2006 2007 2008 2009 2010 Fuel data Total reported fuel consumption [gallons] 1,341,749,279 1,318,070,401 1,207,421,777 982,127,718 772,386,783 934,194,928 Total reported fuel cost [2011$] $2,743,126,593 $3,013,053,236 $2,888,876,874 $3,107,884,964 $1,445,830,162 $2,189,412,000 Reported average price [$/gal] $2.04 $2.29 $2.39 $3.16 $1.87 $2.34 T100 segment data 1 Passenger flights Passengers 2,176,956 2,140,180 2,190,805 2,033,178 1,812,284 1,847,570 Freight [short tons] 17,522 13,990 13,562 11,802 9,402 11,574 Mail [short tons] 1,480 975 615 767 796 821 Total weight [short tons] 236,698 228,984 233,257 215,887 191,427 197,152 Distance [statute miles] 33,408,731 31,103,349 30,721,441 28,703,226 24,719,040 26,182,194 Fuel [gallons] 99,991,701 92,357,776 86,185,405 78,364,628 67,090,646 68,446,811 Fuel cost[$2011] $204,427,084 $211,125,973 $206,207,167 $247,980,222 $125,586,924 $160,414,347 2 Cargo flights Passengers 0 0 0 0 0 0 Freight [short tons] 2,924,469 3,028,992 3,022,954 2,476,231 2,000,499 2,479,659 Mail [short tons] 11,408 8,979 7,766 7,460 5,679 7,742 Total weight [short tons] 2,935,876 3,037,971 3,030,720 2,483,691 2,006,178 2,487,401 Distance [statute miles] 143,972,175 149,297,962 143,534,793 117,699,701 92,561,737 113,779,543 Fuel (gallons) 1,240,246,158 1,225,328,509 1,119,812,037 901,554,281 703,117,216 863,570,472 Fuel cost $2,535,609,499 $2,801,049,191 $2,679,261,842 $2,852,915,098 $1,316,164,517 $2,023,894,047 3 Mixed flights Passengers 21,046 6,847 21,002 33,237 31,854 33,692 Freight [short tons] 1,426 243 1,685 2,635 2,849 2,725 Mail [short tons] 48 25 70 126 183 179 Total weight [short tons] 3,578 952 3,855 6,085 6,217 6,272 Distance [statute miles] 376,726 127,135 433,438 695,035 754,264 746,330 Fuel (gallons) 1,511,421 384,116 1,424,336 2,208,809 2,178,921 2,177,645 Fuel cost $3,090,010 $878,072 $3,407,865 $6,989,644 $4,078,721 $5,103,605 4 Seaplane Passengers 0 0 0 0 0 0 Freight [short tons] ------Mail [short tons] ------Total weight [short tons] ------Distance [statute miles] ------Fuel (gallons) 0 0 0 0 0 0 Fuel cost $0 $0 $0 $0 $0 $0

-C-8-

Appendix D. Glossary of Economic Impact Terms Terms are presented in groups within a logical rather than an alphabetical order

Region defines the geographic area for which impacts are estimated. The region is generally an aggregation of one or more counties. In the case of this Alaska transportation analysis, the region is the state of Alaska.

Sector is a grouping of industries that produce similar products or services. Most economic reporting and models in the U.S. are based on the North American Industrial Classification system (NAIC code). The principle sectors analyzed in this report are the air, water, rail, and trucking transportation sectors.

Impact analysis estimates the impact of dollars from outside the region (“new dollars”) on the region’s economy.

Significance analysis estimates the importance or significance of an industry or activity to a region, usually including spending by both local residents and visitors from outside the region.

Input-output model is a representation of the flows of economic activity between sectors within a region. The model captures what each business or sector must purchase from every other sector in order to produce a dollar’s worth of goods or services. Using such a model, flows of economic activity associated with any change in spending may be traced either forwards (spending generating income which induces further spending) or backwards (industry purchases of fuel that leads refineries to purchase additional inputs – crude oil, utilities, etc.). Multipliers may be derived from an input-output model.

IMPLAN is a micro-computer-based input output modeling system. With IMPLAN, one can estimate 528 sector I-O models for any region consisting of one or more counties. IMPLAN includes procedures for generating multipliers and estimating impacts by applying final demand changes to the model.

Final Demand is the term for sales to final consumers (households or government). Sales between industries are termed intermediate sales. Economic impact analysis generally estimates the regional economic impacts of final demand changes. Household spending is one type of final demand.

Direct effects are the changes in economic activity during the first round of spending. For transportation services this involves the impacts on the transportation industries (businesses selling directly to purchasers) themselves.

Secondary effects are the changes in economic activity from subsequent rounds of re-spending of transportation dollars. There are two types of secondary effects:

Indirect effects are the changes in sales, income or employment within the region in backward- linked industries supplying goods and services to transportation businesses. The increased sales in truck tire supply firms resulting from more shipping services sales is an indirect effect of transportation spending.

-D-1-

Induced effects are the increased sales within the region from household spending of the income earned in transportation services and supporting industries. Employees in transportation services and supporting industries spend the income they earn on housing, utilities, groceries, and other consumer goods and services. This generates sales, income and employment throughout the region’s economy.

Total effects are the sum of direct, indirect, and induced effects.

Multipliers capture the size of the secondary effects in a given region, generally as a ratio of the total change in economic activity in the region relative to the direct change. Multipliers may be expressed as ratios of sales, income or employment, or as ratios of total income or employment changes relative to direct sales. Multipliers express the degree of interdependency between sectors in a region’s economy and therefore vary considerably across regions and sectors.

Type I multipliers measure the direct and indirect effects of a change in economic activity. Unlike Type II or SAM multipliers (discussed below), they do not include induced effects. They capture the inter-industry effects only, i.e., industries buying from local industries.

Type II multipliers capture direct and indirect effects. In addition to the inter-industry effects, the Type II multiplier also takes into account the income and expenditures of households. The household income and the household expenditures are treated as industries. This internalizes the household sector, including induced or household spending, effects.

SAM (IMPLAN Social Accounting Matrix) multipliers are similar to Type II multipliers and use all information about the institutions selected to be included in the predictive model. If only households are included, all information for industries, factors, and households are included.

A sector-specific multiplier gives total changes throughout the economy associated with a unit change in sales in a given sector.

Aggregate multipliers are based on some assumed initial changes in final demand. An aggregate transportation spending multiplier is based on an assumed distribution of transportation spending across economic sectors.

Capture rate is the percentage of spending that accrues to the region’s economy as direct sales or final demand. All transportation spending on services within the region is captured. Generally, however, not all transportation purchases of goods are treated as final demand to the region.

Purchaser prices are the prices paid by the final consumer of a good or service. Producer prices are the prices of goods at the factory or production point.

For manufactured goods the purchaser price = producer price + retail margin + wholesale margin + transportation margin.

For services, the producer and purchaser prices are equivalent.

-D-2-

The retail, wholesale, and transportation margins are the portions of the purchaser price accruing to the retailer, wholesaler, and shipper, respectively. Only the retail margins of many goods purchased by transportation consumers accrue to the local region, if the wholesaler, shipper, and manufacturer lie outside the local area.

Measures of economic activity:

Total Industry Output (TIO): IMPLAN uses input/output accounting to assess the value of production by industry for a calendar year. Output can also be thought of as a value of sales plus or minus inventory.

Sales or output is the dollar volume of a good or service produced or sold

Final Demand = sales to final consumers

Intermediate sales = sales to other industrial sectors

Income is the money earned within the region from production and sales. Total income includes

Wage and salary income, and

Proprietor’s income, rents and profits

Jobs or employment is a measure of the number of jobs required to produce a given volume of sales/production. Jobs are usually not expressed as full-time equivalents, but include part-time and seasonal positions.

Value added is the sum of total income and indirect business taxes. Value added is the most commonly used measure of the contribution of a region to the national economy, as it avoids double counting of intermediate sales and captures only the “value added” by the region to final products.

Passenger revenue ton-mile: One ton of revenue passenger weight (including all baggage) transported one mile. The passenger weight standard for both domestic and international operations is 200 pounds. (BTS5) (BTS6) http://apps.bts.gov/dictionary/search.xml

-D-3-